Structure-based drug design and optimization of mannoside bacterial fimH antagonists

Article abstract of DOI:10.1021/jm100438s

FimH-mediated cellular adhesion to mannosylated proteins is critical in the ability of uropathogenic E. coli (UPEC) to colonize and invade the bladder epithelium during urinary tract infection. We describe the discovery and optimization of potent small-molecule FimH bacterial adhesion antagonists based on α-d-mannose 1-position anomeric glycosides using X-ray structure-guided drug design. Optimized biarylmannosides display low nanomolar binding affinity for FimH in a fluorescence polarization assay and submicromolar cellular activity in a hemagglutination (HA) functional cell assay of bacterial adhesion. X-ray crystallography demonstrates that the biphenyl moiety makes several key interactions with the outer surface of FimH including π-π interactions with Tyr-48 and an H-bonding electrostatic interaction with the Arg-98/Glu-50 salt bridge. Dimeric analogues linked through the biaryl ring show an impressive 8-fold increase in potency relative to monomeric matched pairs and represent the most potent FimH antagonists identified to date. The FimH antagonists described herein hold great potential for development as novel therapeutics for the effective treatment of urinary tract infections.

Full text of DOI:10.1021/jm100438s

J. Med. Chem. 2010, 53, 4779–4792 4779
DOI: 10.1021/jm100438s
Structure-Based Drug Design and Optimization of Mannoside Bacterial FimH Antagonists
Zhenfu Han,†, Jerome S. Pinkner,‡, Bradley Ford,§, Robert Obermann,† William Nolan,† Scott A. Wildman,† Doug Hobbs,†
Tom Ellenberger,† Corinne K. Cusumano,‡ Scott J. Hultgren,*,‡ and James W. Janetka*,†
†Department of Biochemistry and Molecular Biophysics, ‡Department of Molecular Microbiology, Center for Women’s Infectious Disease
Research, and Department of Pathology and Immunology, Washington University School of Medicine, 660 S. Euclid Avenue, St. Louis,
§
Missouri 63110. These authors contributed equally to the work.
Received April 9, 2010
FimH-mediated cellular adhesion to mannosylated proteins is critical in the ability of uropathogenic
E. coli (UPEC) to colonize and invade the bladder epithelium during urinary tract infection. We describe
the discovery and optimization of potent small-molecule FimH bacterial adhesion antagonists based on
R-D-mannose 1-position anomeric glycosides using X-ray structure-guided drug design. Optimized
biarylmannosides display low nanomolar binding affinity for FimH in a fluorescence polarization assay
and submicromolar cellular activity in a hemagglutination (HA) functional cell assay of bacterial
adhesion. X-ray crystallography demonstrates that the biphenyl moiety makes several key interactions
with the outer surface of FimH including π-π interactions with Tyr-48 and an H-bonding electrostatic
interaction with the Arg-98/Glu-50 salt bridge. Dimeric analogues linked through the biaryl ring show
an impressive 8-fold increase in potency relative to monomeric matched pairs and represent the most
potent FimH antagonists identified to date. The FimH antagonists described herein hold great potential
for development as novel therapeutics for the effective treatment of urinary tract infections.
Introduction
rationale for developing novel FimH-binding drugs for the
treatment of UTIs.5 We and others have recently reported the
discovery that R-D-mannosides and glycoconjugate dendrimers
thereof bind with high affinity to the bacterial lectin FimH6 and
prevent hemagglutination of red blood cells mediated by cross-
linking of surface epitopes containing mannose. Ligand bind-
ing affinities have been obtained using diverse quantitative
binding assays and confirmed with the X-ray crystal struc-
ture of various mannosides, such as R-D-butylmannoside
(ButylRman), bound to FimH.6a
Long chain alkyl- and arylmannosides display the highest
affinity of the monovalent ligands likely due to increased
hydrophobic interactions with Ile-52 and two tyrosine resi-
dues, Tyr-48 and Tyr-137, lining the hydrophobic rim of the
binding pocket. A separate crystal structure of the FimH re-
ceptor binding domain boundto oligomannose-3 was recently
solved6c that reveals an “open gate” between Tyr-48 and Tyr-
137 into which oligomannose-3 inserts, adopting a conforma-
tion in which the second mannose residue interacts with
Tyr-48. Interestingly, the latter conformation is different from
that seen in monomannose-bound forms of FimH. This
conformational flexibility of the tyrosine residues seen in these
X-ray structures provides a rationale for designing arylman-
nosides with increased binding affinity relative to alkylman-
nosides by introducing additional hydrophobic and ring
stacking interactions with Tyr-48 or Tyr-137.
FimH is the adhesive component of the type 1 pilus, which
is a member of the chaperone/usher pathway (CUPa) family
of pilus gene clusters. CUP pili function in colonization, in
biofilm formation, and for gaining entry into host cells.1 FimH
is assembled into the tips of type 1 pili, which are essential
virulence factors for the establishment of Escherichia coli
urinary tract infections (UTIs). FimH mediates the coloniza-
tion and invasion of the bladder epithelium.2 After invasion,
UPEC are expelled back out of the bladder cell in a Tlr4
dependent process2c or they escape into the cytoplasm where
they rapidly replicate into intracellular bacterial communities
(IBCs).3 IBC formation is transient and allows the rapid
expansion of a single invaded bacterium into tens of thousands
of bacteria aggregated into a biofilm-like mass in a single
bladder cell. Upon dispersal of UPEC from the IBC biomass,
they are capable of spreading to neighboring cells to repeat the
process, thus further increasing the bacterial load in the
bladder.4 Blocking FimH binding to mannosylated proteins
with FimH antibodies or small molecules is sufficient to prevent
bacterial entry and infection, thus providing the therapeutic
*To whom correspondence should be addressed. For J.W.J.: phone,
314-362-0509; fax, 314-362-7183; e-mail, janetkaj@wustl.edu. For
S.J.H.: phone, 314-362-7059; fax, 314-362-1998; e-mail, hultgren@
biocrim.wustl.edu.
aAbbreviations: UTI, urinary tract infection; UPEC, uropathogenic
E. coli; SAR, structure-activity relationship; HA, hemagglutination;
HAI, HA inhibition; FAM, fluoroscein amidite; SPR, surface plasmon
resonance; CUP, chaperone/usher pathway; IBCs, intracellular bacterial
communities; ButylRman, R-D-butylmannoside; MethylRman, R-D-
methylmannoside; MeUmbRman, 4-methylumbellferyl-R-D-manno-
side; HeptylRman, R-D-heptylmannoside; PhenylRman, R-D-phenyl-
mannoside; PD, pharmacodynamic; PK, pharmacokinetic.
With the exception of one recent communication,6k all
efforts described have only been focused on glycoconjugate
dendrimers. Although synergistic avidity or a “cluster effect”6b
can be realized with the latter agents, multivalent ligands
are predicted to be poorly permeable to the GI tract and
likely not amenable for oral dosing because of their large
r
2010 American Chemical Society
Published on Web 05/27/2010
pubs.acs.org/jmc

4780 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12
Han et al.
Scheme 1a
a Reagents and conditions: (a) R0OH, BF3OEt2, CH2Cl2, reflux; (b) (i) NaOMe, MeOH; (ii) Hþ exchange resin; (c) R0OH, BF3OEt2, CH2Cl2, 0 to
25 °C; (d) H2, 10% Pd/C, EtOH, EtOAc.
Table 1. SAR of Simple Aryl Substitution Mannoside Library
Figure 1. Benzylmannosides.
compd
R
HAI titer (μM) compd
R
HAI titer (μM)
This assay was preferred to a simple FimH binding assay
for screening and developing SAR, since it assesses the com-
pound’s ability to prevent bacterially mediated adhesion
directly in a cellular assay. As shown in Table 1, we found
the general trend that 2- and 3-substitution was optimal for
potency relative to 4-substitution in most examples with the
exception of the acylanilines 3q-s in which this trend was
dramatically reversed. Ortho-substituted chlorophenyl 3g,
cyanophenyl 3m, and meta-substituted methyl ester 3j all
showed an impressive greater than 5-fold improvement in
potency relative to parent phenylmannoside 3a. Interestingly,
carboxylic acid 3l lost 10-fold potency relative to matched pair
methyl ester 3j showing an HAI titer of only 60 μM. Incor-
poration of an additional methyl ester in the 5-position of 3j,
as with compound 4, resulted in a relatively large 3-fold enhan-
cement in potency. Benzylic analogues 5a and 5b (Figure 1),
which have different conformational space and flexibility rela-
tive to direct phenyl substitution of the anomeric oxygen, as
with matched pairs 3a and 3b, show a 2-fold decrease (HAI
titer of 60 μM) in potency.
3a
H
30
31
16
32
4
3m 2-CN
6
23
30
8
3b 4-NO2
3c 3-NO2
3d 4-NH2
3e 3-Me
3n 3-CN
3o 4-CN
3p 4-OMe
3q 2-NHAc
3r 3-NHAc
3s 4-NHAc
3t 3-CONH2
3u 4-CONH2
3v 3-OAc
125
12
8
3f 4-Me
3g 2-Cl
8
4
3h 3-Cl
3i 4-Cl
8
16
15
16
30
2
32
6
3j 3-CO2Me
3k 4-CO2Me
3l 3-CO2H
8
60
3w 4-CH2CO2Me
4
3, 5-CO2Me
molecular weight and high polarity. Therefore, we were
interested in rationally improving both binding affinity and
cellular potency by optimization of monovalent FimH man-
noside antagonists for the clinical treatment of urinary tract
infections.
Results and Discussion
Upon examination of the X-ray structures of ButylRman
and oligomannose-3 coupled with docking studies with man-
noside diester 4 bound to FimH, we were encouraged that fur-
ther improvements in binding affinity could be achieved by
addition of a second aryl or aliphatic ring system in anticipa-
tion of introducing either additional hydrophobic or π-π
stacking interactions with Tyr-48 and/or Tyr-137 as seen from
the directionalityofthe butyl sidechainseen inthe ButylRman
structure and position of the second mannose residue in the
oligomannose-3 bound conformation.
To this end, we explored analogues with an additional ring
system directly attached or fused ring to the parent arylman-
nosides of Table 1. We discovered that a variety of ring
systems were tolerated (Table 2) relative to the previously
described coumarin analogue 4-methylumbellferyl-R-D-man-
noside 6 (MeUmbRman).7 The most attractive inhibitor in
this series of compounds was the 40-biphenyl derivative 8e
Arylmannoside inhibitors of hemagglutination were first
reported over 2 decades ago,7 but surprisingly only limited
structure-activity relationships (SAR) of monovalent aryl-
mannosides are known,6k,7 so we designed and synthesized a
diverse compound set based on aryl substitution of R-D-
phenylmannoside (PhenylRman). As shown in Scheme 1, R-
D-mannoside derivatives were prepared using traditional
Lewis acid mediated glycosidation.8 Reaction of acylated R-
D-(þ)-mannose 1a with a variety of phenols and BF3OEt2
resulted in exclusive formation of the R-isomer mannosides 2.
Subsequent deacylation with NaOMe in methanol gave the
desired arylmannosides 3-8 in good yield. Biological activity
against FimH was evaluated using a guinea pig red blood cell-
based HA assay9 in which HA inhibition (HAI) titers are
defined as the concentration of the compound that results in
>90% HA inhibition.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12 4781
Scheme 2a
Table 2. SAR of Multiring System Analogues
a Reagents and conditions: (a) R0OH, BF3OEt2, CH2Cl2, reflux;
(d) NaOMe, MeOH.
reported in the structure of R-D-mannose,2g while the two
aromatic rings of the biphenyl moiety exist in a nonplanar
conformation allowing for π-stacking and hydrophobic inter-
actions with Tyr-48 and other residues encompassing the
exterior hydrophobic cleft of FimH. The π-stacking inter-
action with Tyr 48 occurs on the opposite side of the
phenyl ring from that engaged by oligomannose-3, which
inserts itself into the open “tyrosine gate” formed by Tyr-48
and Tyr-137. Thus, 8e engages the “tyrosine gate” in a way
that results in alteration of the extended FimH binding
pocket through closure of the tyrosine gate upon rotation
of Tyr-48. This binding mode places the ester of 8e within
H-bonding distance to a salt bridge formed between Arg-98
and Glu-50 (Figure 2). It is noteworthy that the HAI titer
for MethylRman is >1 mM, making compound 8e greater
than 1000-fold more potent from the addition of the biphenyl
ester.
The latter observation led us to investigate replacing the
mannose with alternative sugars or mimics, since the biphenyl
portion alone provides a significant contribution to biological
activity presumably from tighter binding to FimH. We deci-
ded to start by replacing the mannose portion of 8e with
glucose, since all chiral centers are identical except for the
3-hydroxyl group adjacent to the anomeric center, being of
R stereochemistry in mannose and S in glucose (Scheme 2).
Standard Lewis acid mediated glycosidation gave isomeric
mixtures at the anomeric center favoring the undesired β-isomer,
conversely to the clean formation of the R-isomer seen for the
mannosides. Correspondingly, reaction of benzoyl protected
R-D-glucose 9 with phenol 10 using BF3OEt2 followed deben-
zoylation yielded both the R-D-glucoside 11a and β-D-gluco-
side 11b in a 3:7 ratio. Unfortunately, neither glucoside isomer
showed any activity in the hemagglutination assay even up to
2.5 mM. Therefore, modification of a single chiral hydroxyl
group stereocenter on the mannoside is sufficient to sig-
nificantly decrease biological activity, presumably resulting
from the inability to tightly bind FimH. Upon analysis of
the specific interactions realized from R-D-mannose bound
to FimH,2g this hydroxyl group is involved in a H-bonding
interaction with the N-terminal residue of FimH and a
water molecule contained inside the binding pocket. In
any event, this finding demonstrates the exquisite specifi-
city of FimH for mannose epitopes, which has enabled
UPEC bacteria to exclusively recognize mannose-presenting
cells.
Figure 2. X-ray structure of mannoside 8e (PDB code 3MCY) with
electron density (mesh) calculated with 2Fo - Fc coefficients,
contoured at 1σ. Interaction with the Arg-98-Glu-50 salt bridge
(dashes), π-π stacking with Tyr-48, and hydrophobic interaction
with Tyr-137 are shown. Surface electrostatic potential of FimH,
calculated with APBS,10 is displayed such that pure blue and red
would be þ4kT/e and -4kT/e, respectively.
bearing a methyl ester off the meta position of the second aryl
ring and showing a 2-fold improvement relative to 6 having an
HAI titer of 1 μM. In order to determine the source of this
potency enhancement and aid in further improvements, we
obtained a high-resolution X-ray crystal structure of8e bound
to FimH. Shown in Figure 2, inhibitor 8e binds in a very
complementary “lock and key” fashion to the FimH binding
pocket. The mannose ring is making conserved interactions
with the mannose-binding pocket similar to those previously
Next, we returned our attention to optimizing the biaryl
portion of the mannoside by undertaking an extensive SAR
evaluation of the second ring through a modified Topliss-
type11 evaluation of substituents. We designed this focused set

4782 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12
Han et al.
Scheme 3a
a Reagents and conditions: (a) RPh-B(OR)2, Pd(Ph3P)4, dioxane/water (4:1), Cs2CO3, 80 °C; (b) NaOMe, MeOH; (c) H2NMe, EtOH; (d) NaOH (aq),
MeOH.
Table 3. SAR of Substituted Biphenylmannosides
compd
R1
HAI titer (μM)
compd
R1
HAI titer (μM)
14a
14b
14c
14d
14e
14f
14g
14h
14i
3-CONHMe
3-NHAc
3-CO2H
2-CN
1
2
14o
14p
14q
14r
14s
14t
2-NHSO2Me
3-NHSO2Me
2-SO2NHMe
3-SO2NHMe
3-NO2
12
2
4
4
2
1
3-CN
4-CN
1
0.5
2
8
4-NO2
2-CH2OH
3-CH2OH
4-CH2OH
2-OH
16
3
14u
14v
14w
14x
14y
15a
15b
15c
2-CONH2
3-CONH2
2-CF3
6
2
6
8
14j
8
3-CF3
3-F
2
14k
14l
3-OH
4-OH
4
6
6
3, 5-CO2Me
3, 5-CONHMe
3, 5- CO2H
0.15
0.37
2
14m
14n
2-OMe
3-OMe
1
1.5
of compounds with some bias toward improving interactions
with Arg-98 and Glu-50 but also were interested in elucidating
the importance of ring electronic properties as a means to
improve the stacking interaction with Tyr-48. A few initial
analogues were prepared using the synthetic route outlined
previously in Scheme 1, but we developed a more convergent
synthesis based on Suzuki coupling of aryl bromide inter-
mediate 12 as shown in Scheme 3. By employment of standard
Suzuki conditions by reaction of arylboronic acids or esters
with Pd(Ph3P)4 and Cs2CO3, 12 was converted to protected
biphenylmannosides 13 in good yield. The acylated biaryl-
mannosides 13 were then deprotected as beforeto generatethe
final target compound 14 or 15a displayed in Table 3. Diester
15a was further functionalized to dimethylamide 15b by re-
action with dimethylamine or converted to diacid 15c by basic
hydrolysis in methanol. Upon evaluation of HA titers, the
matched pair meta-substituted methylamide 14a was equipo-
tent to ester 8e while reverse amide 14b and free acid 14c
displayed moderately lower potencies of 2 and 4 μM, respec-
tively. The latter result was surprising, since these modifica-
tions were designed to introduce an electrostatic interaction
with Arg-98 and to disrupt the Arg-98/Glu-50 salt bridge. On
the other hand, sulfonamide 14r has equivalent potency to
carboxamide 14a, providing further evidence that a hydrogen
bond acceptor is required for optimal potency presumably
from interaction with Arg-98. Upon analysis of ortho, meta,
and para matched pairs, it was obvious that para substitution
was least preferred for activity while meta substitution was
preferred to ortho substitution. This preference for meta
substitution was most pronounced with the methyl alcohols
14g and 14h where the ortho analogue 14g was over 5-fold less
potent with an HAI titer of only 16 μM. We surmise that this
large substituent results in an increased rotation of the two
rings out of the plane, causing a nonproductive alignment for
interactions with Tyr-48. Although a majority of the analogues
tested contained electron withdrawing groups, in general elec-
tron donating groups such as the methyl alcohols 14g-i and
phenols 14j-l were less active in the HA titer assay. The most

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12 4783
Scheme 4a
a Reagents and conditions: (a) BF3OEt2, CH2Cl2, reflux; (b) (i) NaOMe, MeOH; (c) 19, Et3N, DMF.
Table 4. Binding Affinity of Selected Mannosides in Fluorescence Polarization Assay
compd
FP binding EC50 (μM)
HAI titer (μM)
compd
FP binding EC50 (μM)
HAI titer (μM)
6 (MeUmbRman)
8e
0.044
0.052
0.059
0.064
0.07
2
HeptylRman
4
14p
0.089
0.091
0.097
0.099
0.114
0.122
0.160
0.162
0.187
0.245
15
2
1
8f
3k
4
2
8
14c
3a
4
14b
14a
15a
14v
3j
2
30
8
0.075
0.077
0.078
0.084
0.087
1
3s
0.15
2
3r
3n
12
23
60
125
6
0.37
5a
ButylRman
15b
potent of the monosubstituted mannosides is meta nitro deri-
vative 14m with an HAI titer of 0.5 μM. One possible explana-
tion for the improved potency of 14m is that the partial negative
charge on the nitro oxygen atoms allow for an optimized
H-bond acceptor-donor interaction with Arg-98 coupled with
the increased electron withdrawing ability of the nitro group
possibly enhancing stacking interactions with Tyr-48.
deprotection of intermediate 18 to give free amine 19. Sub-
sequent reaction of 5(6)FAM-OSu with 19 using triethyla-
mine in DMF gave the desired fluorescently tagged FAM-
mannoside 20.
The KD for FAM mannoside 20 was measured to be 0.17 μM,
while the HAI titer was much lower at 125 μM. We evaluated 20
mannoside ligands with structural diversity and activities in the
HA cell assay ranging from 150 nM to 125 μM for their ability
to competitively inhibit binding of 5(6)-FAM 20. Interestingly,
we found that all potently competed for FimH binding having
EC50 less than 0.25 μM but having no clear correlation of EC50
with the activity found in the HA cell assay (Table 4). Although
there was over a 60-fold range in cellular potency, there was only
about a 5-fold range in binding affinity (Table 5). In addition,
there was also no linear correlation of data with the two assays.
Simple alkylmannosides have the highest drop in cell activity, as
demonstrated with ButylRman having the most dramatic differ-
ence with a 510-fold drop in cell activity relative to binding.
Interestingly, the optimized biphenylmannoside series
shows a modest correlation of binding affinity having a much
smaller drop in cell activity relative to binding. The diester 15a
and diamide 15b show the best correlation with only a 2-fold
and 4-fold difference in cell potency vs binding. It is important
to note that even though these two analogues are the most
potent in the HA assay, they have about 2-fold less binding
affinity in the FP assay than the tightest binding mannosides.
While it is unclear why there is no correlation of binding to
activity in the cell, there is a moderate correlation when com-
pounds with binned affinities are compared. However, the
latter observation cannot explain the nonlinear increase in cell
activity of 15a and 15b relative to the monosubstituted
biphenylmannosides. Perhaps differential binding kinetics
including varied ON rates and OFF rates of mannoside analo-
gues to FimH can provide a plausible explanation for the
Perhaps the most exciting finding from this study was that
addition of another ester or amide substituted in the other
meta position such as diester 15a and dimethylamide 15b
resulted in the two most potent mannoside inhibitors of FimH
withactivities of 150 and 370 nM, respectively. With respect to
diester 15a this constitutes a 7-fold improvement in activity
relative to mono ester 8e. The diamide 15b, while slightly less
potent, has much improved solubility relative to diester 15a.
We hypothesize that the addition of the second ester or amide
serves a 2-fold purpose for improving activity. First, we
propose that the electron withdrawing group results in less
electron density of the aryl ring, thus improving π-π stacking
interactions with Tyr-48. Second, the monoester (or amide)
analogue can presumably access conformations in which the
aryl ring isnot in proximitytoArg-98 whereas the metadiester
(or diamide) analogue likely exists only in conformations where
the ester or amide resides in proximity to Arg-98 resulting in less
entropic loss upon binding and a lower energy bound con-
formation resulting in increased binding affinity and potency.
In order to more accurately determine the relative contribu-
tions of binding affinity to FimH versus other compound
properties to the potency seen in the cellular HA assay, we
developed a high-throughput fluorescence polarization/ani-
sotropy competitive binding assay using a fluorescent-labeled
mannoside ligand shown in Scheme 4. Synthesis was achieved
as before using Lewis acid mediated glycosylation with pro-
tected amino alcohol 17 followed by dual Fmoc and acyl

4784 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12
Table 5. Correlation of HAI Titer and FimH Binding Data
Han et al.
latter. It is worth mentioning that the previously reported KD
values6a obtained through surface plasmon resonance (SPR)
for 6 (MeUmbRman), ButylRman, and HeptylRman are 20,
151, and 5 nM, respectively. While the result for ButylRman is
similar, the FP assay suggests that MeUmbRman is the most
potent (44 nM), being 2-fold more potent than HeptylRman
(89 nM). Considering the lack of correlation between binding
affinity and biological activity in the cell plus the variability
seen comparing two separate binding assays, the HA assay
coupled with the use of structural information is a preferred
route for medicinal chemistry optimization of mannoside
FimH ligands. Nonetheless, based on the HAI titer cellular
data, mannoside 15a is the most potent FimH antagonist re-
ported to date and represents an excellent lead candidate for
further optimization of the monovalent mannosides toward
a novel preclinical candidate with tremendous therapeutic
potential for treating urinary tract infections. Several compo-
unds show pharmacodynamic (PD) activity in novel murine
models of UTI and are currently being optimized for in vivo
efficacy and pharmacokinetic (PK) properties. The results of
these studies will be reported shortly in a future manuscript.
The development of multivalent and dendrimeric manno-
sides has been the major focus of previous work on FimH
antagonists because simple mannosides, while having respect-
able binding affinity to FimH, show poor cellular activity in
the HA assay. It has been demonstrated that multivalent man-
nosides are effective at increasing avidity through a phenom-
enon termed “cluster effect” resulting in an overall increased
binding affinity when calculated per mannoside monomer.
Therefore, we were curious if our improved monovalent
mannosides 15a and 15b would produce a similar increase in
avidity or cluster effect previously reported for dendrimeric
mannosides. For synthetic simplicity, we chose to utilize a
model system based on monoamide 14a. A symmetrical diva-
lent mannoside was designed by connecting two monomers
via an amide based linker to the two biphenyl rings (Scheme 5).
Accordingly, dimeric inhibitors based on 14a were synthesized
by coupling carboxylic acid 13c to either diamine 2-(2-ami-
noethoxy)ethanamine or 2-[2-(2-aminoethoxy)ethoxy]ethana-
mine, yielding diamide 21 or diamide 22, respectively, using
standard HATU coupling conditions with Hunig’s base in
DMF. The shorter chain analogue 21 showed an HAI titer of
0.75 μM, thus showing no noticeable improvement relative to
14a, while the longer ethylene glycol linked diamide 22 dis-
played an almost 8-fold increase in activity with an impressive
HAI titer of 130 nM. This constitutes a 4-fold increase in
avidity relative to the expected potency of 500 nM based on
two monomeric units of 14a. In addition to increased potency,
22 has much improved solubility relative to 21, making it an
ideal starting point for further optimization. Although they are
less likely to be effective as oral agents compared with the
monovalent mannosides, the divalent mannosides are very
useful chemical research tools and can potentially be devel-
oped into topical or intravenously dosed antibacterial agents.
We are currently exploring optimized divalent analogues based
on disubstituted 15b which are predicted to have low nano-
molar activity in the HA titer assay based on our current know-
ledge and understanding of SAR.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12 4785
Scheme 5a
a Reagents and conditions: (a) HATU, [H2N(CH2)2]2O, DIPEA, DMF; (b) HATU, [H2N(CH2)2OCH2]2, DIPEA, DMF.
Conclusions
primarily focused on improvements in pharmacokinetics and
in vivo efficacy in animal models, which will be reported in
another communication.
We have designed a series of potent small-molecule FimH
antagonists using the weak inhibitor R-D-phenylmannoside as
an initial starting point for X-ray structure-guided optimiza-
tion. Addition of substituents with increased hydrophobicity
resulted in potency enhancements most pronounced using
aromatic groups. Uponseveral rounds of SAR evaluation and
optimization, we discovered that 40-biaryl groups substituted
on the meta position with H-bond acceptors such as esters
or amides yield the most potent analogues. The structural
basis for the enhanced potency results from both an optimal
π-stacking arrangement of the biaryl moiety with Tyr-48 and
an H-bonding interaction between an H-bond acceptor such
as an amide carbonyl with Arg-98. We have also designed
dimers derived from these mannosides, which show a 4-fold
increase in cellular potency relative to that expected from
2 monomeric units. This improvement in avidity can be attri-
buted to the ability of the dimer to bind two separate FimH
molecules either from adjacent bacterial pili or on another
bacterium. This issimilar tothe “cluster effect” of dendrimeric
mannosides described recently by others. Both the optimized
monomeric and dimeric mannosides described herein repre-
sent the most potent FimH antagonists reported to date and
are very attractive leads as novel therapeutics for the treat-
ment of urinary tract infections in which there exists a large
unmet medicalneed. FimH mediatedrecognitionof uroplakin
receptors through the mannose binding epitope and subse-
quent invasion of UPEC into bladder cells is only one of a vast
number of adhesion mechanisms involving a diverse range of
carbohydrate epitopes which both bacterial and viral patho-
gens utilize for colonization.12 The ability to design and opti-
mize carbohydrate-derived or glycomimetic antagonists utili-
zing X-ray structure-based design strategies, such as described
for FimH in this report, is instrumental for identifying other
antagonists of carbohydrate-protein interactions and anti-
virulence therapeutics. It is important to mention that several
carbohydrate-derived or glycomimetic drugs either are being
evaluated in clinical trials or have reached the marketplace for
the treatment of various diseases.13 In fact, 1,6-bis[3-(3-car-
boxymethylphenyl)-4-(2-R-D-mannopyranosyloxy)phenyl]-
hexane (TBC1269), a dimeric biarylmannoside antagonist14
of E, P-, and L-selectin, is currently undergoing phase IIa
evaluation for the treatment of asthma and psoriasis. We are
currently pursuing the further optimization and preclinical
evaluation of the biarylmannoside class of FimH antagonists
Experimental Section
General. 1H NMR spectra were obtained on a Varian 300 MHz
NMR instrument. The chemical shifts were reported as δ ppm
relative toTMS, using residual solvent peak as the reference unless
otherwise noted. The following abbreviations wereused to express
the multiplicities: s=singlet; d=doublet; t=triplet; q=quartet;
m=multiplet; br=broad. High-performance liquid chromatog-
raphy (HPLC) was carried out on a GILSON GX-281 using
Waters C18 5 μM, 4.6 mm ꢀ 50 mm and Waters Prep C18 5 μM,
19 mm ꢀ 150 mm reverse phase columns. Mass spectra (MS) were
obtained on a Waters MicromassZQ. All reactions were mon-
itored by thin layer chromatography (TLC) carried out on Merck
silica gel plates (0.25 mm thick, 60F254), visualized by using UV
(254 nm) or dyes such as KMnO4, p-anisaldehyde, and CAM
(ceric ammonium molybdate). Silica gel chromatography was
carried out on an ISCO system using ISCO prepacked silica gel
columns (12-330 g sizes). All compounds used for biological
assays are at least of 95% purity based on HPLC analytical
results monitored with 220 and 254 nm wavelengths, unless
otherwise noted. All reagents and solvents were purchased from
commercially available sources and used as received except
resorcinol monoacetate, which was purified prior to use. Phenyl-
R-D-mannopyranoside, p-nitrophenyl-R-D-mannopyranoside,
p-aminophenyl-R-D-mannopyranoside, and 4-methylumbelli-
feryl-R-D-mannopyranoside were purchased from Sigma.
Synthesis of Mannosides. General Procedure for the Prepara-
tion of Mannosides Using r-D-Mannose Pentaacetate: Methyl
3-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)tet-
rahydropyran-2-yl]oxyphenyl]benzoate (8e). Methyl 3-[4-[(2R,
3S,4S,5R,6R)-3,4,5-Triacetoxy-6-(acetoxymethyl)tetrahydropyran-
2-yl]oxyphenyl]benzoate. Under nitrogen atmosphere, at 0 °C
boron trifluoride diethyl etherate (0.128 g, 0.90 mmol) was added
dropwise into a solution of R-D-mannose pentaacetate (0.120 g,
0.3 mmol) and methyl 3-(4-hydroxyphenyl)benzoate (0.140 g,
0.6 mmol) in 6 mL of CH2Cl2. After a few minutes the mixture
was heated to reflux and was kept stirring for more than 36 h. The
reaction was then quenched with water and extracted with
CH2Cl2. The CH2Cl2 layer was collected, dried with Na2SO4,
and concentrated in vacuo. The resulting residue was purified by
silica gel chromatography with hexane/ethyl acetate combinations
as eluent, giving rise to methyl 3-[4-[(2R,3S,4S,5R,6R)-3,4,5-tria-
cetoxy-6-(acetoxymethyl)tetrahydropyran-2-yl]oxyphenyl]benzo-
ate (0.136 g) in 81% yield. 1H NMR (300 MHz, CDCl3) δppm 8.23
(m, 1H), 7.99(m, 1H), 7.74(m, 1H), 7.57(m, 2H), 7.50(t, J=7.8Hz,
1H),7.18(m,2H),5.59(dd, J=3.6, 9.9 Hz, 1H), 5.58 (d, J=1.5Hz,

4786 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12
Han et al.
1H), 5.48 (dd, J=2.1, 3.3 Hz, 1H), 5.39 (t, J=10.2 Hz, 1H), 4.30
(dd, J= 5.4, 12.3 Hz, 1H), 4.06-4.15 (m, 2 H), 3.95 (s, 3H), 2.15
(s, 3H), 2.06 (s, 3H), 2.05 (s, 3H), 2.04 (s, 3H). MS (ESI): found
[M þ H]þ, 559.1.
(300 MHz, deuterium oxide) δ ppm 3.61-3.69 (m, 1 H), 3.70-
3.80 (m, 3 H), 3.90 (s, 3 H), 4.06 (dd, J=3.60, 9.30 Hz, 1 H), 4.17
(dd, J=3.30, 1.80 Hz, 1 H), 5.72 (d, J=1.80 Hz, 1 H), 7.20-7.26
(m, 2 H), 7.98 - 8.04 (m, 2 H). MS (ESI): found [M þ Na]þ,
337.4.
Methyl 3-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxy-
methyl)tetrahydropyran-2-yl]oxyphenyl]benzoate (8e). Methyl 3-
[4-[3,4,5-triacetoxy-6-(acetoxymethyl)tetrahydropyran-2-yl]oxy-
phenyl]benzoate (0.120 g) was stirred in 6 mL of methanol with a
catalytic amount of sodium methoxide (0.02 M) at room tem-
perature overnight. Hþ exchange resin (DOWEX 50WX4-100)
was added to neutralize the mixture. The resin was filtered off and
the filtrate was concentrated and then dried in vacuo, giving rise
to pure product 8e (0.084 g) in quantitative yield. 1H NMR (300
MHz, CD3OD) δ ppm 8.21 (m, 1H), 7.95 (m, 1H), 7.83 (m, 1H),
7.59 (m, 2H), 7.53 (m, 1H), 7.23 (m, 2H), 5.55 (d, J=1.8 Hz, 1H),
4.03 (dd, J=1.8, 3.3 Hz, 1H), 3.91-3.96 (m, 4H), 3.70-3.82 (m,
3H), 3.59-3.65 (m, 1H). MS (ESI): found [M þ H]þ, 391.1.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-(3-nitrophenoxy)tet-
rahydropyran-3,4,5-triol (3c). 3c was prepared using the same
procedure as for 8e. Yield: 46%. 1H NMR (300 MHz, methanol-
d4) δ ppm 3.51-3.60 (m, 1 H), 3.67-3.81 (m, 3 H), 3.87-3.95
(m, 1 H), 4.05 (dd, J=3.30, 1.65 Hz, 1 H), 5.61 (d, J=1.65 Hz,
1 H), 7.50-7.60 (m, 2 H), 7.85-7.94 (m, 1 H), 7.97 (dt, J=2.75,
1.10 Hz, 1 H). MS (ESI): found [2M þ H]þ, 603.1.
2-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxybenzonitrile (3m). 3m was prepared
using the same procedure as for 8e and was further purified by
silica gel chromatography with methylene chloride and metha-
nol combination as eluent. Yield: 55%. 1H NMR (300 MHz,
methanol-d4) δ ppm 3.51-3.63 (m, 1 H), 3.65-3.85 (m, 3 H),
3.98 (dd, J = 9.48, 3.43 Hz, 1 H), 4.10 (dd, J = 3.30, 1.92 Hz,
1 H), 5.66 (d, J = 1.65 Hz, 1 H), 7.15 (td, J = 7.55, 1.10 Hz,
1 H), 7.48 (d, J=8.24 Hz, 1 H), 7.56-7.70 (m, 2 H). MS (ESI):
found [M þ H]þ, 281.8.
3-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)tet-
rahydropyran-2-yl]oxybenzonitrile (3n). 3n was prepared using
the same procedure as for 8e. Yield: 64%. 1H NMR (300 MHz,
methanol-d4) δ ppm 3.48-3.60 (m, 1 H), 3.63-3.82 (m, 3 H),
3.89 (dd, J = 9.48, 3.43 Hz, 1 H), 4.02 (dd, J = 3.30, 1.92 Hz,
1 H), 5.55 (d, J=1.37 Hz, 1 H), 7.30-7.57 (m, 4 H). MS (ESI):
found [M þ K]þ, 320.1.
4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)tet-
rahydropyran-2-yl]oxybenzonitrile (3o). 3o was prepared using
the same procedure as for 8e. Yield: 33%. 1H NMR (300 MHz,
methanol-d4) δ ppm 3.44-3.60 (m, 1 H), 3.60-3.82 (m, 3 H),
3.88 (dd, J=9.34, 3.30 Hz, 1 H), 3.96-4.19 (m, 1 H), 5.48-5.74
(m, 1 H), 7.27 (d, J=8.79 Hz, 2 H), 7.68 (d, J=8.79 Hz, 2 H).
MS (ESI): found [2M þ Na]þ, 585.8.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-(3-methylphenoxy)-
tetrahydropyran-3,4,5-triol (3e). 3e was prepared using the same
procedure as for 8e. Yield: 27%. 1H NMR (300 MHz, methanol-
d4) δ ppm 2.30 (s, 3 H), 3.56-3.66 (m, 1 H), 3.68-3.82 (m, 3 H),
3.87-3.95 (m, 1 H), 3.99 (dd, J=3.43, 1.79 Hz, 1 H), 5.45 (d, J=
1.92 Hz, 1 H), 6.79-7.00 (m, 3 H), 7.14 (t, J=7.83 Hz, 1 H). MS
(ESI): found [2M þ H]þ, 540.9.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-(4-methoxyphenoxy)-
tetrahydropyran-3,4,5-triol (3p). 3p was prepared using the same
procedure as for 8e. Yield: 56%. 1H NMR (300 MHz, methanol-
d4) δ ppm 3.61-3.68 (m, 1 H), 3.69-3.83 (m, 6 H), 3.85-3.92
(m, 1 H), 3.99 (dd, J=3.30, 1.92 Hz, 1 H), 5.34 (d, J=1.92 Hz,
1 H), 6.78-6.91 (m, 2 H), 6.96-7.12 (m, 2 H). MS (ESI): found
[M þ Na]þ, 308.7.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-(4-methylphenoxy)-
tetrahydropyran-3,4,5-triol (3f). 3f was prepared using the same
procedure as for 8e. Yield: 44%. 1H NMR (300 MHz, methanol-
d4) δ ppm 2.26 (s, 3 H), 3.57-3.66 (m, 1 H), 3.68-3.82 (m, 3 H),
3.86-3.95 (m, 1 H), 3.99 (dd, J=3.30, 1.92 Hz, 1 H), 5.41 (d, J=
1.92 Hz, 1 H), 6.95-7.04 (m, 2 H), 7.05-7.13 (m, 2 H). MS
(ESI): found [2M þ H]þ, 541.4.
N-[2-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]acetamide (3q). 3q was prepa-
red using the same procedure as for 8e. Yield: 42%. H NMR
1
(2R,3S,4S,5S,6R)-2-(2-Chlorophenoxy)-6-(hydroxymethyl)-
tetrahydropyran-3,4,5-triol (3g). 3g was prepared using the same
procedure as for 8e. Yield: 29%. 1H NMR (300 MHz, methanol-
d4) δ ppm 3.60-3.69 (m, 1 H), 3.70-3.81 (m, 3 H), 3.98 (dd, J=
9.20, 3.43 Hz, 1 H), 4.10 (dd, J=3.30, 1.92 Hz, 1 H), 5.53 (d, J=
1.65 Hz, 1 H), 7.00 (td, J = 7.69, 1.37 Hz, 1 H), 7.25 (ddd, J =
8.24, 7.42, 1.65 Hz, 1 H), 7.32-7.41 (m, 2 H). MS (ESI): found
[M þ Na]þ, 312.7.
(300 MHz, methanol-d4) δ ppm 2.17 (s, 3 H), 3.56-3.67 (m,
1 H), 3.67-3.83 (m, 3 H), 3.93 (dd, J=9.07, 3.30 Hz, 1 H), 4.09
(br. s., 1 H), 5.46 (s, 1 H), 6.92-7.07 (m, 1 H), 7.15 (t, J =
7.14 Hz, 1 H), 7.34 (d, J=7.97 Hz, 1 H), 7.53-7.75 (m, 1 H). MS
(ESI): found [M þ Na]þ, 335.9.
N-[3-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]acetamide (3r). 3r was prepared
using the same procedure as for 8e. Yield: 79%. 1H NMR
(300 MHz, methanol-d4) δ ppm 2.11 (s, 3 H), 3.54-3.65 (m, 1
H), 3.67-3.82 (m, 3 H), 3.86-3.95 (m, 1 H), 4.00 (dd, J=3.30,
1.92 Hz, 1 H), 5.47 (d, J = 1.37 Hz, 1 H), 6.79-6.96 (m, 1 H),
7.05-7.31 (m, 2 H), 7.43 (t, J=2.06 Hz, 1 H). MS (ESI): found
[M þ H]þ, 314.3.
(2R,3S,4S,5S,6R)-2-(3-Chlorophenoxy)-6-(hydroxymethyl)-
tetrahydropyran-3,4,5-triol (3h). 3h was prepared using the same
procedure as for 8e. Yield: 58%. 1H NMR (300 MHz, methanol-
d4) δ ppm 3.51-3.61 (m, 1 H), 3.67-3.81 (m, 3 H), 3.84-3.92
(m, 1 H), 4.00 (dd, J=2.88, 2.06 Hz, 1 H), 5.48 (d, J=1.65 Hz,
1 H), 6.97-7.09 (m, 2 H), 7.12-7.18 (m, 1 H), 7.21-7.30 (m,
1 H). MS (ESI): found [M þ K]þ, 328.6.
N-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]acetamide (3s). 3s was prepared
using the same procedure as for 8e. Yield: 38%. 1H NMR
(300 MHz, methanol-d4) δ ppm 2.10 (s, 3 H), 3.53-3.65 (m,
1 H), 3.67-3.82 (m, 3 H), 3.83-3.94 (m, 1 H), 3.99 (dd, J=3.43,
1.79 Hz, 1 H), 5.42 (d, J = 1.65 Hz, 1 H), 6.98-7.18 (m, 2 H),
7.36-7.57 (m, 2 H). MS (ESI): found [M þ H]þ, 313.7.
3-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)tet-
rahydropyran-2-yl]oxybenzamide (3t). 3t was prepared using the
(2R,3S,4S,5S,6R)-2-(4-Chlorophenoxy)-6-(hydroxymethyl)-
tetrahydropyran-3,4,5-triol (3i). 3i was prepared using the same
procedure as for 8e. Yield: 65%. 1H NMR (300 MHz, methanol-
d4) δ ppm 3.52-3.62 (m, 1 H), 3.66-3.81 (m, 3 H), 3.83-3.91
(m, 1 H), 3.99 (dd, J=3.43, 1.79 Hz, 1 H), 5.45 (d, J=1.65 Hz,
1 H), 7.06-7.17 (m, 2 H), 7.22-7.32 (m, 2 H). MS (ESI): found
[2M þ Na]þ, 603.0.
Methyl 3-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxym-
ethyl)tetrahydropyran-2-yl]oxybenzoate (3j). 3j was prepared
using the same procedure as for 8e. Yield: 75%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.52-3.64 (m, 1 H), 3.65-3.82
(m, 3 H), 3.84-3.98 (m, 4 H), 4.02 (dd, J= 3.30, 1.92 Hz, 1 H),
5.54 (d, J = 1.65 Hz, 1 H), 7.31-7.51 (m, 2 H), 7.62-7.81 (m,
2 H). MS (ESI): found [M þ Na]þ, 336.8.
1
same procedure as for 8e. Yield: 24%. H NMR (300 MHz,
methanol-d4) δ ppm 3.55-3.64 (m, 1 H), 3.65-3.82 (m, 3 H),
3.87-3.95 (m, 1 H), 4.03 (dd, J=3.30, 1.92 Hz, 1 H), 5.55 (d, J=
1.92 Hz, 1 H), 7.26-7.33 (m, 1 H), 7.39 (t, J=7.83 Hz, 1 H), 7.53
(dt, J=7.55, 1.17 Hz, 1 H), 7.58-7.65 (m, 1 H). MS (ESI): found
[2M þ H]þ, 597.9.
Methyl 4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxym-
ethyl)tetrahydropyran-2-yl]oxybenzoate (3k). 3k was prepa-
red using the same procedure as for 8e. Yield: 60%. 1H NMR
4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)tet-
rahydropyran-2-yl]oxybenzamide (3u). 3u was prepared using the
same procedure as for 8e. Yield: 38%. H NMR (300 MHz,
1

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12 4787
methanol-d4) δ ppm 3.49-3.60 (m, 1 H), 3.65-3.81 (m, 3 H),
3.86-3.94 (m, 1 H), 4.02 (dd, J=3.43, 1.79 Hz, 1 H), 5.57 (d, J=
1.65 Hz, 1 H), 7.12-7.28 (m, 2 H), 7.79-7.93 (m, 2 H). MS
(ESI): found [M þ H]þ, 300.9.
nol-d4) δ ppm 3.54 (td, J = 4.74, 2.88 Hz, 1 H), 3.63-3.78 (m,
4 H), 3.82 (t, J=2.20 Hz, 1 H), 5.40 (d, J=1.92 Hz, 1 H), 7.06-
7.15 (m, 1 H), 7.26-7.35 (m, 3 H), 7.35-7.44 (m, 3 H), 7.44-
7.51 (m, 2 H). MS (ESI): found [2M þ H]þ, 665.1.
Methyl 2-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxy-
methyl)tetrahydropyran-2-yl]oxyphenyl]acetate (3w). 3w was pre-
pared using the same procedure as for 8e. Yield: 74%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.54-3.64 (m, 3 H), 3.65-3.82
(m, 6 H), 3.86-3.94 (m, 1 H), 3.99 (dd, J= 3.30, 1.92 Hz, 1 H),
5.45 (d, J = 1.65 Hz, 1 H), 6.98-7.14 (m, 2 H), 7.20 (d, J =
8.79 Hz, 2 H). MS (ESI): found [M þ H]þ, 328.5.
Methyl 4-[3-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxy-
methyl)tetrahydropyran-2-yl]oxyphenyl]benzoate (8b). 8b was
prepared using the same procedure as for 8e. Yield: 77%. H
1
NMR (300 MHz, methanol-d4) δ ppm 3.60-3.84 (m, 4 H), 3.86-
3.98(m, 4H), 4.04(dd, J=3.43, 1.79 Hz, 1 H), 5.56 (d, J=1.92Hz,
1 H), 7.17 (ddd, J=8.04, 2.40, 1.10 Hz, 1 H), 7.28-7.54 (m, 3 H),
7.66-7.85 (m, 2 H), 8.00-8.22 (m, 2 H). MS (ESI): found
[M þ H]þ, 391.6.
Dimethyl 5-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxy-
methyl)tetrahydropyran-2-yl]oxybenzene-1,3-dicarboxylate (4).
4 was prepared using the same procedure as for 8e. Yield:
75%. 1H NMR (300 MHz, methanol-d4) δ ppm 3.49-3.62
(m, 1 H), 3.66-3.85 (m, 3 H), 3.86-4.01 (m, 7 H), 4.02-4.11 (m,
1 H), 5.62 (s, 1 H), 7.94 (d, J=1.10 Hz, 2 H), 8.27 (s, 1 H). MS
(ESI): found [M þ H]þ, 373.0.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-(4-phenylphenoxy)-
tetrahydropyran-3,4,5-triol (8c). 8c was prepared using the same
procedure as for 8e. Yield: 73%. 1H NMR (300 MHz, methanol-
d4) δ ppm 3.57-3.67 (m, 1 H), 3.69-3.84 (m, 3 H), 3.87-3.97
(m, 1 H), 4.03 (dd, J=3.57, 1.92 Hz, 1 H), 5.53 (d, J=1.92 Hz, 1
H), 7.15-7.23 (m, 2 H), 7.24-7.32 (m, 1 H), 7.35-7.45 (m, 2 H),
7.49-7.60 (m, 4 H). MS (ESI): found [2M þ H]þ, 664.9.
Phenyl-[3-[(2R,3S,4S,5S,6R)-3,4,5-trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]methanone (8d). 8d was pre-
pared using the same procedure as for 8e. Yield: 83%. 1H
NMR (300 MHz, methanol-d4) δ ppm 3.56-3.66 (m, 1 H), 3.70-
3.85 (m, 3 H), 3.88-3.98 (m, 1 H), 4.05 (dd, J = 3.43, 1.79 Hz,
1 H), 5.56 (d, J=1.65 Hz, 1 H), 7.33-7.46 (m, 3 H), 7.46-7.57
(m, 3 H), 7.57-7.67 (m, 1 H), 7.70-7.83 (m, 2 H). MS (ESI):
found [M þ H]þ, 361.1.
(2R,3S,4S,5S,6R)-2-Benzyloxy-6-(hydroxymethyl)tetrahydro-
pyran-3,4,5-triol (5a). 5a was prepared using the same procedure
as for 8e. Yield: 52%. H NMR (300 MHz, deuterium oxide)
1
δ ppm 3.59-3.88 (m, 5 H), 3.92 (dd, J=3.30, 1.65 Hz, 1 H), 4.55
(d, J=11.54 Hz, 1 H), 4.75 (d, J=11.54 Hz, 1 H), 4.95 (d, J=
1.65 Hz, 1 H), 7.28-7.56 (m, 5 H). MS (ESI): found [M þ Na]þ,
293.0.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[(4-nitrophenyl)met-
hoxy]tetrahydropyran-3,4,5-triol (5b). 5b was prepared using the
same procedure as for 8e. Yield: 76%. H NMR (300 MHz,
1
Methyl 4-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxy-
methyl)tetrahydropyran-2-yl]oxyphenyl]benzoate (8f). 8f was pre-
pared using the same procedure as for 8e. Yield: 40%. 1H NMR
(300 MHz, methanol-d4/chloroform-d1) δ ppm 3.56-3.69 (m, 1
H), 3.71-3.86 (m, 3 H), 3.88-3.99 (m, 4 H), 4.04 (dd, J =3.30,
1.92 Hz, 1 H), 5.54 (d, J = 1.65 Hz, 1 H), 7.03-7.23 (m, 2 H),
7.53-7.58 (m, 2 H), 7.62 (d, J=8.52 Hz, 2 H), 8.04 (d, J=8.52
Hz, 2 H). MS (ESI): found [M þ K]þ, 429.5.
methanol-d4) δ ppm 3.53-3.65 (m, 2 H), 3.65-3.80 (m, 2 H),
3.82-3.88 (m, 1 H), 3.89 (dd, J=3.43, 1.79 Hz, 1 H), 4.67 (d, J=
13.19 Hz, 1 H), 4.92 (d, J=13.19 Hz, 1 H), 7.60-7.65 (m, 2 H),
8.15-8.33 (m, 2 H). MS (ESI): found [M þ Na]þ, 337.6.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-(2-naphthyloxy)tetra-
hydropyran-3,4,5-triol (7b). 7b was prepared using the same
1
procedure as for 8e. Yield: 50%. H NMR (300 MHz, metha-
N-[3-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]phenyl]acetamide (14b). 14b was
prepared using the same procedure as for 8e. Yield: 50%. H
nol-d4) δ ppm 3.60-3.69 (m, 1 H), 3.70-3.87 (m, 3 H), 3.97 (dd,
J=9.34, 3.57 Hz, 1 H), 4.07 (dd, J=3.43, 1.79 Hz, 1 H), 5.63 (d,
J=1.65 Hz, 1 H), 7.24 (dd, J=8.79, 2.47 Hz, 1 H), 7.30-7.38
(m, 1 H), 7.39-7.48 (m, 1 H), 7.57 (d, J = 2.47 Hz, 1 H),
7.71-7.90 (m, 3 H). MS (ESI): found [M þ H]þ, 307.0.
1
NMR (300 MHz, methanol-d4) δ ppm 2.14 (s, 3 H), 3.63 (ddd, J=
7.28, 4.94, 2.33 Hz, 1 H), 3.68-3.84 (m, 3 H), 3.88-3.98 (m, 1 H),
4.04 (dd, J=3.43, 1.79 Hz, 1 H), 5.53 (d, J=1.65 Hz, 1 H), 7.09-
7.23 (m, 2 H), 7.24-7.39 (m, 2 H), 7.41-7.62 (m, 3 H), 7.78 (t, J=
1.65 Hz, 1 H). MS (ESI): found [M þ H]þ, 390.2.
6-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)tet-
rahydropyran-2-yl]oxy-3a,7a-dihydro-3H-1,3-benzoxazol-2-one
(7c). 7c was prepared using the same procedure as for 8e and was
further purified by silica gel chromatography with methylene
Dimethyl 5-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydro-
xymethyl)tetrahydropyran-2-yl]oxyphenyl]benzene-1,3-dicarboxy-
late (15a). 15a was prepared using the same procedure as for 8e
and was further purified by silica gel chromatography with
chloroform and methanol combination as eluent. Yield: 53%.
1H NMR (300 MHz, methanol-d4) δ ppm 3.57-3.66 (m, 1 H),
3.68-3.84 (m, 3 H), 3.93 (dd, J = 9.34, 3.30 Hz, 1 H), 3.97 (s,
6 H), 4.04 (dd, J=3.30, 1.92 Hz, 1 H), 5.56 (d, J=1.65 Hz, 1 H),
7.21-7.36 (m, 2 H), 7.59-7.72 (m, 2 H), 8.42 (d, J=1.65 Hz, 2 H),
8.55 (t, J=1.51 Hz, 1 H). MS (ESI): found [M þ H]þ, 449.6.
Procedure for the Preparation of Mannosides Using 2,3,4,6-
Tetra-O-benzyl-1-acetyl-r-D-mannopyranose:3 [3-[(2R,3S,4S,5S,6R)-
3,4,5-Trihydroxy-6-(hydroxymethyl)tetrahydropyran-2-yl]oxyp-
henyl]acetate (3v). [3-[(2R,3S,4S,5R,6R)-3,4,5-Tribenzyloxy-6-
(benzyloxymethyl)tetrahydropyran-2-yl]oxyphenyl]acetate. Under
nitrogen atmosphere, at 0 °C boron trifluoride diethyl etherate
(0.051 g, 0.36 mmol) was added dropwise into the solution of 2,3,4,6-
tetra-O-benzyl-1-acetyl-R-D-mannopyranose (0.106 g, 0.18 mmol)
and resorcinol monoacetate (0.055 g, 0.36 mmol) in 7 mL of CH2Cl2.
The mixture was stirred at 0 °C and monitored by TLC. The
reaction was then quenched with water and extracted with CH2Cl2.
The CH2Cl2 layer was collected, dried with Na2SO4, and concen-
trated in vacuo. The resulting residue was purified by silica gel
chromography with hexane/ethyl acetate combinations as eluent to
give [3-[(2R,3S,4S,5R,6R)-3,4,5-tribenzyloxy-6-(benzyloxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]acetate (0.093 g) in 75% yield. 1H
NMR (300 MHz, CDCl3) δ ppm 7.15-7.40 (m, 21H), 6.88-6.91
1
chloride and methanol combination as eluent. Yield: 32%. H
NMR (300 MHz, methanol-d4) δ ppm 3.59-3.67 (m, 1 H),
3.67-3.83 (m, 3 H), 3.84-3.91 (m, 1 H), 4.01 (dd, J=3.30, 1.92
Hz, 1 H), 5.40 (d, J=1.65 Hz, 1 H), 6.89-7.04 (m, 2 H), 7.13 (d,
J =2.20 Hz, 1 H). MS (ESI): found [M þ Na]þ, 336.0.
Methyl 5-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxybenzothiophene-2-carboxylate (7d). 7d
was prepared using the same procedure as for 8e. Yield: 66%. 1H
NMR (300 MHz, methanol-d4) δ ppm 3.59-3.67 (m, 1 H),
3.68-3.84 (m, 3 H), 3.89-3.97 (m, 4 H), 4.05 (dd, J = 3.30,
1.92 Hz, 1 H), 5.55 (d, J=1.65 Hz, 1 H), 7.27 (dd, J=8.79, 2.47
Hz, 1 H), 7.72 (d, J=2.47 Hz, 1 H), 7.81 (d, J=8.79 Hz, 1 H),
8.01 (s, 1 H). MS (ESI): found [M þ H]þ, 371.6.
Methyl 6-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxy-
methyl)tetrahydropyran-2-yl]oxybenzothiophene-2-carboxylate
(7e). 7e was prepared using the same procedure as for 8e. Yield:
50%. 1H NMR (300 MHz, methanol-d4) δ ppm 3.55-3.64 (m,
1 H), 3.67-3.81 (m, 3 H), 3.84-3.98 (m, 4 H), 4.05 (dd, J=3.43,
1.79 Hz, 1 H), 5.58 (d, J = 1.65 Hz, 1 H), 7.19 (dd, J = 8.79,
2.20 Hz, 1 H), 7.74 (d, J = 2.20 Hz, 1 H), 7.85 (d, J = 8.79 Hz,
1 H), 8.02 (s, 1 H). MS (ESI): found [M þ H]þ, 371.5.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-(2-phenylphenoxy)-
tetrahydropyran-3,4,5-triol (8a). 8a was prepared using the same
procedure as for 8e and was further purified by HPLC (C18,
15 mm ꢀ 150 mm column; eluent acetonitrile/water (0.1%
TFA)). Yield: 60%. Purity: 85%. 1H NMR (300 MHz, metha-

4788 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12
Han et al.
(m, 1H), 6.81(t, J= 2.4 Hz, 1H), 6.73-6.76 (m, 1H), 5.58 (d, J=
1.8 Hz, 1H), 4.90 (d, J=10.8 Hz, 1H), 4.78 (s, 2H), 4.64-4.69 (m,
3H), 4.52 (d, J=10.8 Hz, 1H), 4.45 (d, J=12.3 Hz, 1H), 4.05-4.18
(m, 2H), 3.94 (t, J=2.4 Hz, 1H), 3.77-3.85 (m, 2H), 3.64-3.70 (m,
1H), 2.27 (s, 3H). MS (ESI): found [M þ Na]þ, 697.2.
(0.016 g) in 71% yield. Glucoside R-anomer 11a: 1H NMR
(300 MHz, methanol-d4) δ ppm 3.40-3.50 (m, 1 H), 3.56-3.63
(m, 1 H), 3.64-3.81 (m, 3 H), 3.89 (dd, J=9.75, 8.93 Hz, 1 H),
3.94 (s, 3 H), 5.55 (d, J = 3.57 Hz, 1 H), 7.23-7.36 (m, 2 H),
7.48-7.69 (m, 3 H), 7.83 (ddd, J=7.83, 1.92, 1.24 Hz, 1 H), 7.96
(dt, J=7.76, 1.48 Hz, 1 H), 8.21 (t, J=1.65 Hz, 1 H). MS (ESI):
found [M þ H]þ, 391.4. Glucoside β-anomer 11b: 1H NMR (300
MHz, methanol-d4) δ ppm 3.36-3.56 (m, 4 H), 3.72 (dd, J =
11.95, 5.36 Hz, 1 H), 3.85-4.01 (m, 4 H), 4.93-5.04 (m, 1 H),
7.12-7.30 (m, 2 H), 7.45-7.65 (m, 3 H), 7.75-7.87 (m, 1 H),
7.96 (dq, J=7.69, 0.92 Hz, 1 H), 8.15-8.27 (m, 1 H). MS (ESI):
found [M þ H]þ, 391.7.
[3-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]acetate (3v). Under hydrogen
atmosphere [3-[3,4,5-tribenzyloxy-6-(benzyloxymethyl)tetrahydro-
pyran-2-yl]oxyphenyl]acetate (0.085 g, 0.13 mmol) was stirred
with Pd/C(10 wt %) (0.132 g, 0.063 mmol) in ethanol (6 mL) and
ethyl acetate (6 mL). The reaction was monitored by TLC until it
went into completion. The mixture was filtered through a Celite
plug. The filtrate was concentrated and then dried in vacuo,
furnishing pure 3v (0.040 g) in quantitative yield. 1H NMR (300
MHz, CD3OD) δ ppm 7.30 (t, J=8.1 Hz, 1H), 7.00 (m, 1H), 6.90
(t, J=2.1 Hz, 1H), 6.76 (m, 1H), 5.47 (d, J=1.8 Hz, 1H), 4.00 (dd,
J=1.8, 3.3 Hz, 1H), 3.88 (dd, J=3.3, 9.3 Hz, 1H), 3.68-3.79 (m,
3H), 3.54-3.61 (m, 1H). MS (ESI): found [M þ H]þ, 314.9.
5-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxy-3H-benzofuran-2-one (7a). 7a prepared
in the same way as 3v. Yield: 59%. Purity: 91%. 1H NMR
(300 MHz, D2O) δ ppm 7.19 (m, 1H), 7.11 (m, 2H), 5.52 (d, J=
1.5 Hz, 1H), 4.15-4.18 (m, 1H), 3.97-4.05 (m, 1H), 3.68-3.88
(m, 6H). MS (ESI): found [M þ H]þ, 313.0.
General Procedure for the Preparation of Biphenylmannoside
Derivatives through Suzuki Coupling Reaction: 3-[4-[(2R,3S,
4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)tetrahydropyran-
2-yl]oxyphenyl]benzonitrile (14e). [(2R,3S,4S,5R,6R)-3,4,5-Tri-
acetoxy-6-[4-(3-cyanophenyl)phenoxy]tetrahydropyran-2-yl]methyl
Acetate. Under nitrogen atmosphere, a mixture of acetyl-pro-
tected 4-bromophenylmannoside 12 (0.101 g, 0.2 mmol), m-cya-
nophenylboronic acid (0.044 g, 0.3 mmol), cesium carbonate
(0.196 g, 0.6 mmol), and tetrakis(triphenylphosphine)palladium
(0.023 g, 0.02 mmol) in dioxane/water (5 mL/1 mL) was heated
at 80 °C with stirring for 1 h. The solvent was removed and the
resulting residue was purified by silica gel chromography with
hexane/ethyl acetate combinations as eluent to give [3,4,5-
triacetoxy-6-[4-(3-cyanophenyl)phenoxy]tetrahydropyran-2-yl]-
Procedure for the Preparation of Glucoside Analogues: Methyl
3-[4-[(2R,3R,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzoate (r-Anomer, 11a) and
Methyl 3-[4-[(2S,3R,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzoate (β-Anomer, 11b). Methyl
3-[4-[(2R,3R,4S,5R,6R)-3,4,5-Tribenzoyloxy-6-(benzoyloxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzoate (r-Anomer) and Methyl
3-[4-[(2S,3R,4S,5R,6R)-3,4,5-Tribenzoyloxy-6-(benzoyloxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzoate (β-Anomer). Under nitro-
gen atmosphere, at 0 °C boron trifluoride diethyl etherate (0.212 g,
1.5 mmol) was added dropwise into a solution of glucose penta-
benzoate 9 (0.350 g, 0.5 mmol) and methyl 3-(4-hydroxyphe-
nyl)benzoate 10 (0.228 g, 1.0 mmol) in 10 mL of CH2Cl2. After a
few minutes the mixture was heated to reflux and was kept stirring
for more than 36 h. The reaction was then quenched with water
and extracted with CH2Cl2. The CH2Cl2 layer was collected, dried
with Na2SO4, then concentrated in vacuo. The residue was purified
by HPLC (C18, 15 mm ꢀ 150 mm column; eluent acetonitrile/
water (0.1% TFA)) to give benzoyl-protected glucosides, methyl
3-[4-[(2R,3R,4S,5R,6R)-3,4,5-tribenzoyloxy-6-(benzoyloxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzoate (R-anomer) (0.026 g)
and methyl 3-[4-[(2S,3R,4S,5R,6R)-3,4,5-tribenzoyloxy-6-(benzoy-
loxymethyl)tetrahydropyran-2-yl]oxyphenyl]benzoate (β-anomer)
(0.045 g) totally in 18% yield. Benzoyl-protected glucoside
R-anomer: 1H NMR (300 MHz, acetonitrile-d3) δ ppm 3.86-3.94
(m, 3 H), 4.44-4.56 (m, 2 H), 4.64-4.75 (m, 1 H), 5.66 (dd, J =
10.16, 3.57 Hz, 1 H), 5.79 (t, J=9.89 Hz, 1 H), 6.12 (d, J=3.85 Hz,
1H), 6.32(t, J=9.89Hz, 1H), 7.27-7.33 (m, 2 H), 7.34-7.47 (m, 8
H), 7.49-7.63 (m, 7 H), 7.77 (ddd, J = 7.69, 1.92, 1.10 Hz, 1 H),
7.83-7.89 (m, 2 H), 7.90-7.99 (m, 7 H), 8.11-8.19 (m, 1 H). MS
(ESI): found [M þ Na]þ, 829.5. Benzoyl-protected glucoside β-
anomer: 1H NMR (300 MHz, acetonitrile-d3) δ ppm 3.91 (s, 3 H),
4.50-4.69 (m, 3 H), 5.68-5.88 (m, 3 H), 5.98-6.14 (t, J=9.6 Hz,
1 H), 7.06-7.17 (m, 2 H), 7.32-7.70 (m, 15 H), 7.73-7.87 (m, 3 H),
7.88-8.01 (m, 5 H), 8.02-8.11 (m, 2 H), 8.14 (t, J=1.79 Hz, 1 H).
MS (ESI): found [M þ Na]þ, 829.8.
1
methyl acetate (0.080 g) in 76% yield. H NMR (300 MHz,
CDCl3) δ ppm 7.82 (t, J = 1.5 Hz, 1H), 7.76 (dt, J = 7.69,
1.65 Hz, 1H), 7.58-7.65 (m, 1H), 7.47-7.57 (m, 3H), 7.16-7.24
(m, 2H), 5.55-5.65 (m, 2H), 5.47 (dd, J = 3.57, 1.92 Hz, 1H),
5.39 (t, J=9.9 Hz, 1H), 4.25-4.34 (m, 1H), 4.04-4.17 (m, 2H),
2.22 (s, 3H), 2.06 (s, 3H), 2.05 (s, 3H), 2.04 (s, 3H). MS (ESI):
found [M þ Na]þ, 548.7.
3-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzonitrile (14e). [3,4,5-Triace-
toxy-6-[4-(3-cyanophenyl)phenoxy]tetrahydropyran-2-yl]methyl
acetate (0.075 g) was stirred in 6 mL of methanol with a catalytic
amount of sodium methoxide (0.02 M) at room tempera-
ture overnight. Hþ exchange resin (DOWEX 50WX4-100) was
added to neutralize the mixture. The resin was filtered off and
the filtrate was concentrated andthen dried in vacuo, giving rise
to pure product 14e (0.045 g) in 88% yield. 1H NMR (300 MHz,
CD3OD) δ ppm 3.55-3.65 (m, 1H), 3.67-3.83 (m, 3H),
3.84-3.99 (m, 1H), 4.03 (dd, J = 3.43, 1.79 Hz, 1H), 5.55 (J =
1.65 Hz, 1H), 7.16-7.33 (m, 2H), 7.54-7.75 (m, 4H), 7.83-8.01
(m, 2H). MS (ESI): found [M þ H]þ, 358.3.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-(3-pyridyl)phenoxy]-
tetrahydropyran-3,4,5-triol (8g). 8g was prepared using the same
procedure as for 14e except that Hþ exchange resin was not used.
The product was further purified by silica gel chromatography
with methylene chloride and methanol combination with 3%
1
(v/v) ammonia aqueous as eluent. Yield: 65%. H NMR (300
MHz, methanol-d4) δ ppm 3.56-3.67 (m, 1 H), 3.68-3.85 (m,
3 H), 3.87-3.97 (m, 1 H), 4.04 (dd, J = 3.30, 1.92 Hz,
1 H), 5.56 (d, J = 1.92 Hz, 1 H), 7.21-7.32 (m, 2 H), 7.44-
7.55 (m, 1 H), 7.56-7.69 (m, 2 H), 8.00-8.14 (m, 1 H), 8.47 (dd,
J = 4.94, 1.37 Hz, 1 H), 8.72-8.82 (m, 1 H). MS (ESI): found
[M þ H]þ, 334.4.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-(4-pyridyl)phenoxy]-
tetrahydropyran-3,4,5-triol (8h). 8h was prepared using the same
procedure as for 14e except that Hþ exchange resin was not used.
The product was further purified by silica gel chromatography
with methylene chloride and methanol combination with 3%
(v/v) ammonia aqueous as eluent. Yield: 12%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.60 (ddd, J = 7.28, 4.94, 2.33
Hz, 1 H), 3.67-3.82 (m, 3 H), 3.87-3.97 (m, 1 H), 4.04 (dd, J=
3.43, 1.79 Hz, 1 H), 5.57 (d, J = 1.65 Hz, 1 H), 7.19-7.34 (m,
2 H), 7.61-7.82 (m, 4 H), 8.54 (d, J=5.22 Hz, 2 H). MS (ESI):
found [M þ H]þ, 334.5.
Methyl 3-[4-[(2R,3R,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxy-
methyl)tetrahydropyran-2-yl]oxyphenyl]benzoate (r-Anomer, 11a)
and Methyl 3-[4-[(2S,3R,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxy-
methyl)tetrahydropyran-2-yl]oxyphenyl]benzoate (β-Anomer, 11b).
Deprotection of the benzoyl-protected glucoside R-anomer and
benzoyl-protected glucoside β-anomer with a catalytic amount
of sodium methoxide in methanol was done in the same way as
for 8e, and subsequent purification by HPLC (C18, 15 mm ꢀ
150 mm column; eluent acetonitrile/water (0.1% TFA)) furn-
ished glucosides R-anomer 11a (0.009 g) and β-anomer 11b

Article
2-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12 4789
using the same procedure as for 14e. Yield: 75%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.60-3.68 (m, 1 H), 3.69-3.83
(m, 6 H), 3.89-3.97 (m, 1 H), 4.03 (dd, J=3.30, 1.92 Hz, 1 H),
5.52 (d, J = 1.92 Hz, 1 H), 6.94-7.07 (m, 2 H), 7.10-7.17 (m,
2 H), 7.21-7.33 (m, 2 H), 7.37-7.45 (m, 2 H). MS (ESI): found
[M þ H]þ, 363.2.
tetrahydropyran-2-yl]oxyphenyl]benzonitrile (14d). 14d was pre-
pared using the same procedure as for 14e. Yield: 12%. It came
from the minor product of the Suzuki coupling with 2-cyano-
phenylboronic acid. 1H NMR (300 MHz, methanol-d4) δ ppm
3.57-3.67 (m, 1 H), 3.69-3.85 (m, 3 H), 3.88-3.98 (m, 1 H),
4.04 (dd, J = 3.43, 1.79 Hz, 1 H), 5.57 (d, J = 1.65 Hz, 1 H),
7.21-7.32 (m, 2 H), 7.45-7.60 (m, 4 H), 7.66-7.76 (m, 1 H),
7.80 (dd, J = 7.69, 1.37 Hz, 1 H). MS (ESI): found [M þ H]þ,
358.4.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-(3-methoxyphenyl)-
phenoxy]tetrahydropyran-3,4,5-triol (14n). 14n was prepared
using the same procedure as for 14e. Yield: 54%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.50-3.59 (m, 1 H), 3.60-3.78
(m, 6 H), 3.85 (dd, J = 9.34, 3.30 Hz, 1 H), 3.95 (dd, J = 3.30,
1.92 Hz, 1 H), 5.45 (d, J = 1.65 Hz, 1 H), 6.79 (ddd, J = 8.17,
2.54, 0.82 Hz, 1 H), 6.96-7.16 (m, 4 H), 7.18-7.31 (m, 1 H),
7.38-7.54 (m, 2 H). MS (ESI): found [M þ H]þ, 363.2.
N-[2-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]phenyl]methanesulfonamide (14o).
14o was prepared using the same procedure as for 14e. Yield:
4-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzonitrile (14f). 14f was pre-
pared using the same procedure as for 14e. Yield: 67%. 1H
NMR (300 MHz, methanol-d4) δ ppm 3.52-3.67 (m, 1 H),
3.67-3.84 (m, 3 H), 3.87-3.98 (m, 1 H), 4.03 (dd, J=3.43, 1.79
Hz, 1 H), 5.56 (d, J = 1.65 Hz, 1 H), 7.17-7.33 (m, 2 H),
7.56-7.70 (m, 2 H), 7.72-7.85 (m, 4 H). MS (ESI): found
[M þ H]þ, 358.3.
1
63%. H NMR (300 MHz, methanol-d4) δ ppm 2.71 (s, 3 H),
3.57-3.66 (m, 1 H), 3.67-3.83 (m, 3 H), 3.93 (dd, J = 9.34,
3.57 Hz, 1 H), 4.04 (dd, J = 3.43, 1.79 Hz, 1 H), 5.56 (d, J =
1.65 Hz, 1 H), 7.18-7.26 (m, 2 H), 7.26-7.41 (m, 5 H), 7.48 (d,
J =7.69 Hz, 1 H). MS (ESI): found [M þ H]þ, 426.6.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-[2-(hydroxymethyl)-
phenyl]phenoxy]tetrahydropyran-3,4,5-triol (14g). 14g was pre-
pared using the same procedure as for 14e. Yield: 79%. 1H
NMR (300 MHz, methanol-d4) δ ppm 3.59-3.69 (m, 1 H),
3.70-3.83 (m, 3 H), 3.89-3.98 (m, 1 H), 4.04 (dd, J = 3.57,
1.92 Hz, 1 H), 4.51 (s, 2 H), 5.53 (d, J=1.65 Hz, 1 H), 7.14-7.25
(m, 3 H), 7.27-7.43 (m, 4 H), 7.55 (dd, J=7.55, 1.51 Hz, 1 H).
MS (ESI): found [M þ Li]þ, 365.4.
N-[3-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]phenyl]methanesulfonamide (14p).
14p was prepared using the same procedure as for 14e. Yield:
1
52%. H NMR (300 MHz, methanol-d4) δ ppm 2.98 (s, 3 H),
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-[3-(hydroxymethyl)-
phenyl]phenoxy]tetrahydropyran-3,4,5-triol (14h). 14h was pre-
pared using the same procedure as for 14e. Yield: 70%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.59-3.68 (m, 1 H), 3.69-3.84 (m,
3 H), 3.89-3.97 (m, 1 H), 4.03 (dd, J=3.43, 1.79 Hz, 1 H), 4.66 (s,
2 H), 5.53 (d, J=1.65 Hz, 1 H), 7.15-7.23 (m, 2 H), 7.26-7.33 (m,
1 H), 7.38 (t, J=7.42 Hz, 1 H), 7.44-7.50 (m, 1 H), 7.52-7.61 (m,
3 H). MS (ESI): found [M þ Li]þ, 365.5.
3.56-3.66 (m, 1 H), 3.68-3.83 (m, 3 H), 3.88-3.97 (m, 1 H),
4.03 (dd, J = 3.43, 1.79 Hz, 1 H), 5.53 (d, J = 1.92 Hz, 1 H),
7.17-7.25 (m, 3 H), 7.33-7.42 (m, 2 H), 7.43-7.48 (m, 1 H),
7.50-7.63 (m, 2 H). MS (ESI): found [M þ H]þ, 426.2.
N-Methyl-2-[4-[(2R,3S,4S,5S,6R)-3,4,5-trihydroxy-6-(hydro-
xymethyl)tetrahydropyran-2-yl]oxyphenyl]benzenesulfonamide
(14q). 14q was prepared using the same procedure as for 14e.
Yield: 50%. 1H NMR (300 MHz, methanol-d4) δ ppm 2.36 (s,
3 H), 3.59-3.68 (m, 1 H), 3.68-3.84 (m, 3 H), 3.88-3.96 (m,
1 H), 4.04 (dd, J=3.43, 1.79 Hz, 1 H), 5.56 (d, J=1.92 Hz, 1 H),
7.13-7.21 (m, 2 H), 7.28-7.39 (m, 3 H), 7.49-7.58 (m, 1 H),
7.58-7.68 (m, 1 H), 8.03 (dd, J=7.83, 1.24 Hz, 1 H). MS (ESI):
found [M þ H]þ, 426.1.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-[4-(hydroxymethyl)-
phenyl]phenoxy]tetrahydropyran-3,4,5-triol (14i). 14i was pre-
pared using the same procedure as for 14e and was further
purified by HPLC (C18, 15 mm ꢀ 150 mm column; eluent ace-
tonitrile/water (0.1% TFA)). Yield: 50%. 1H NMR (300 MHz,
methanol-d4) δ ppm 3.58-3.69 (m, 1 H), 3.69-3.83 (m, 3 H),
3.89-3.98 (m, 1 H), 4.03 (dd, J = 3.30, 1.92 Hz, 1 H), 4.63 (s,
2 H), 5.52 (d, J=1.65 Hz, 1 H), 7.14-7.28 (m, 2 H), 7.40 (d, J=
8.52 Hz, 2 H), 7.50-7.66 (m, 4 H). MS (ESI): found [M þ Li]þ,
365.8.
N-Methyl-3-[4-[(2R,3S,4S,5S,6R)-3,4,5-trihydroxy-6-(hydro-
xymethyl)tetrahydropyran-2-yl]oxyphenyl]benzenesulfonamide
(14r). 14r was prepared using the same procedure as for 14e.
Yield: 54%. 1H NMR (300 MHz, methanol-d4) δ ppm 2.55 (s,
3 H), 3.56-3.66 (m, 1 H), 3.68-3.83 (m, 3 H), 3.89-3.97 (m,
1 H), 4.04 (dd, J=3.43, 1.79 Hz, 1 H), 5.55 (d, J=1.92 Hz, 1 H),
7.20-7.30 (m, 2 H), 7.53-7.68 (m, 3 H), 7.77 (dt, J = 8.03,
1.34 Hz, 1 H), 7.82-7.89 (m, 1 H), 8.02 (t, J=1.79 Hz, 1 H). MS
(ESI): found [M þ H]þ, 426.0.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-(2-hydroxyphenyl)-
phenoxy]tetrahydropyran-3,4,5-triol (14j). 14j was prepared
using the same procedure as for 14e. Yield: 74%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.60-3.68 (m, 1 H), 3.69-3.84
(m, 3 H), 3.89-3.97 (m, 1 H), 4.03 (dd, J=3.30, 1.92 Hz, 1 H),
5.52 (d, J = 1.65 Hz, 1 H), 6.83-6.92 (m, 2 H), 7.07-7.18 (m,
3 H), 7.19-7.25 (m, 1 H), 7.43-7.55 (m, 2 H). MS (ESI): found
[M þ H]þ, 349.4.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-(3-nitrophenyl)-
phenoxy]tetrahydropyran-3,4,5-triol (14s). 14s was prepared
using the same procedure as for 14e. Yield: 70%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.61 (ddd, J = 9.75, 4.94,
2.61 Hz, 1 H), 3.68-3.84 (m, 3 H), 3.87-3.98 (m, 1 H), 4.04
(dd, J=3.43, 1.79 Hz, 1 H), 5.56 (d, J=1.92 Hz, 1 H), 7.18-7.35
(m, 2 H), 7.57-7.74 (m, 3 H), 7.92-8.08 (m, 1 H), 8.17 (dd, J=
8.24, 2.20 Hz, 1 H), 8.42 (t, J=2.06 Hz, 1 H). MS (ESI): found
[2M þ H]þ, 755.8.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-(3-hydroxyphenyl)-
phenoxy]tetrahydropyran-3,4,5-triol (14k). 14k was prepared
using the same procedure as for 14e. Yield: 66%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.56-3.67 (m, 1 H), 3.68-3.83
(m, 3 H), 3.88-3.96 (m, 1 H), 4.03 (dd, J=3.43, 1.79 Hz, 1 H),
5.52 (d, J = 1.65 Hz, 1 H), 6.73 (ddd, J = 8.10, 2.33, 1.10 Hz,
1 H), 6.88-7.06 (m, 2 H), 7.07-7.28 (m, 3 H), 7.38-7.60 (m,
2 H). MS (ESI): found [M þ H]þ, 349.3.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-(4-nitrophenyl)-
phenoxy]tetrahydropyran-3,4,5-triol (14t). 14t was prepared
using the same procedure as for 14e. Yield: 60%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.56-3.64 (m, 1 H), 3.68-3.84
(m, 3 H), 3.87-3.97 (m, 1 H), 4.04 (dd, J=3.43, 1.79 Hz, 1 H),
5.57 (d, J=1.65 Hz, 1 H), 7.20-7.33 (m, 2 H), 7.64-7.76 (m, 2
H), 7.79-7.90 (m, 2 H), 8.24-8.38 (m, 2 H). MS (ESI): found
[M þ H]þ, 378.5.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-(4-hydroxyphenyl)-
phenoxy]tetrahydropyran-3,4,5-triol (14l). 14l was prepared
using the same procedure as for 14e. Yield: 71%. 1H NMR
(300 MHz, methanol-d4) δ ppm 3.58-3.68 (m, 1 H), 3.69-3.84
(m, 3 H), 3.86-3.97 (m, 1 H), 4.02 (dd, J=3.43, 1.79 Hz, 1 H),
5.49 (d, J = 1.92 Hz, 1 H), 6.79-6.86 (m, 2 H), 7.11-7.19 (m,
2 H), 7.35-7.42 (m, 2 H), 7.43-7.50 (m, 2 H). MS (ESI): found
[2M þ H]þ, 697.6.
2-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzamide (14u). 14u was pre-
pared using the same procedure as for 14e. Yield: 42%. It came
from the major product of the Suzuki coupling with 2-cyano-
phenylboronic acid. 1H NMR (300 MHz, methanol-d4) δ ppm
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-(2-methoxyphenyl)-
phenoxy]tetrahydropyran-3,4,5-triol (14m). 14m was prepared

4790 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12
Han et al.
3.56-3.68 (m, 1 H), 3.68-3.84 (m, 3 H), 3.87-3.97 (m, 1 H),
4.02 (dd, J = 3.30, 1.92 Hz, 1 H), 5.51 (d, J = 1.92 Hz, 1 H),
7.11-7.21 (m, 2 H), 7.33-7.44 (m, 4 H), 7.44-7.57 (m, 2 H). MS
(ESI): found [M þ H]þ, 376.4.
1.65 Hz, 1 H), 7.11-7.20 (m, 2 H), 7.44 (t, J=7.69 Hz, 1 H), 7.52
(m, J = 8.79 Hz, 2 H), 7.73 (d, J = 7.69 Hz, 1 H), 7.89 (d, J =
7.69 Hz, 1 H), 8.09-8.20 (m, 1 H). MS (ESI): found [M þ H]þ,
377.5.
3-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxybenzoic Acid (3l). 31 was prepared
through the hydrolysis of 3j, using the same method as for
15c, in quantitative yield and in pure form. 1H NMR (300 MHz,
methanol-d4) δ ppm 3.52-3.64 (m, 1 H), 3.67-3.83 (m, 3 H),
3.85-3.96 (m, 1 H), 4.03 (dd, J=3.43, 1.79 Hz, 1 H), 5.55 (d, J=
1.65 Hz, 1 H), 7.32-7.46 (m, 2 H), 7.64-7.78 (m, 2 H). MS
(ESI): found [M þ H]þ, 323.1.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-[2-(trifluoromethyl)-
phenyl]phenoxy]tetrahydropyran-3,4,5-triol (14w). 14w was pre-
pared using the same procedure as for 14e. Yield: 67%. 1H
NMR (300 MHz, methanol-d4) δ ppm 3.60-3.70 (m, 1 H),
3.70-3.86 (m, 3 H), 3.93 (dd, J = 9.34, 3.30 Hz, 1 H), 4.04
(dd, J=3.30, 1.92 Hz, 1 H), 5.54 (d, J=1.65 Hz, 1 H), 7.12-7.30
(m, 4 H), 7.34 (d, J=6.87 Hz, 1 H), 7.47-7.58 (m, 1 H), 7.62 (t,
J=7.14 Hz, 1 H), 7.76 (d, J=7.97 Hz, 1 H). MS (ESI): found
[2M þ H]þ, 801.8.
3-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzamide (14v). 8e (0.060 g,
0.150 mmol) was stirred in 15 mL of aqueous ammonia (29 wt
%) at room temperature for 24 h. The solvent was removed and
the residue was purified by HPLC (C18, 15 mm ꢀ 150 mm
column; eluent acetonitrile/water (0.1% TFA)) to give 14v
(0.037 g) in 65% yield. 1H NMR (300 MHz, methanol-d4)
δ ppm 3.58-3.66 (m, 1 H), 3.69-3.82 (m, 3 H), 3.93 (dd, J =
9.30, 3.30 Hz, 1 H), 4.03 (dd, J=3.30, 1.50 Hz, 1 H), 5.54 (d, J=
1.80 Hz, 1 H), 7.18-7.26 (m, 2 H), 7.51 (t, J=7.80 Hz, 1 H), 7.58
- 7.66 (m, 2 H), 7.73-7.83 (m, 2 H), 8.10 (t, J=1.5 Hz, 1 H). MS
(ESI): found [M þ H]þ, 376.3.
(2R,3S,4S,5S,6R)-2-(Hydroxymethyl)-6-[4-[3-(trifluoromethyl)-
phenyl]phenoxy]tetrahydropyran-3,4,5-triol (14x). 14x was pre-
pared using the same procedure as for 14e. Yield: 70%. 1H
NMR (300 MHz, methanol-d4) δ ppm 3.56-3.66 (m, 1 H),
3.67-3.84 (m, 3 H), 3.93 (dd, J=9.34, 3.57 Hz, 1 H), 4.03 (dd,
J=3.43, 1.79 Hz, 1 H), 5.55 (d, J=1.65 Hz, 1 H), 7.17-7.32 (m,
2 H), 7.55-7.66 (m, 4 H), 7.79-7.89 (m, 2 H). MS (ESI): found
[2M þ H]þ, 801.8.
(2R,3S,4S,5S,6R)-2-[4-(3-Fluorophenyl)phenoxy]-6-(hydroxy-
methyl)tetrahydropyran-3,4,5-triol (14y). 14y was prepared
using the same procedure as for 14e. Yield: 70%. H NMR
1
(300 MHz, methanol-d4) δ ppm 3.56-3.67 (m, 1 H), 3.68-
3.85 (m, 3 H), 3.87-3.97 (m, 1 H), 4.03 (dd, J=3.43, 1.79 Hz,
1 H), 5.54 (d, J=1.65 Hz, 1 H), 6.96-7.08 (m, 1 H), 7.14-7.24
(m, 2 H), 7.26-7.35 (m, 1 H), 7.36-7.46 (m, 2 H), 7.51-7.62
(m, 2 H). MS (ESI): found [2M þ H]þ, 701.8.
[(2S,3S,4S,5R,6R)-4,5-Diacetoxy-6-(acetoxymethyl)-2-[4-(9H-
fluoren-9-ylmethoxycarbonylamino)butoxy]tetrahydropyran-3-yl]-
acetate (18). Under nitrogen atmosphere, at 0 °C boron tri-
fluoride diethyl etherate (0.115 g, 0.81 mmol) was added drop-
wise into a solution of R-D-mannose pentaacetate (0.107 g,
0.27 mmol) and 4-(Fmoc-amino)-1-butanol (0.168 g, 0.54 mmol)
in 5 mL of CH2Cl2. After a few minutes the mixture was heated
to reflux and was kept stirring for more than 36 h. The reaction
was then quenched with water and extracted with CH2Cl2. The
CH2Cl2 layer was collected, dried with Na2SO4, and concen-
trated. The resulting residue was purified by silica gel chromo-
graphy with hexane/ethyl acetate combinations as eluent to give
18 (0.106 g) in 60% yield. 1H NMR (300 MHz, CDCl3) δ ppm
1.48-1.79 (m, 4H), 2.01 (s, 3H), 2.05 (s, 3H), 2.11 (s, 3H), 2.17 (s,
3H), 3.25 (m, 2H), 3.50 (m, 1H), 3.73 (m, 1H), 3.91-4.04 (m,
1H), 4.05-4.18 (m, 1H), 4.18-4.37 (m, 2H), 4.41 (d, J=6.87 Hz,
2H), 4.74-4.94 (m, 2H), 5.17-5.41 (m, 3H), 7.29-7.37 (m, 2H),
7.37-7.47 (m, 2H), 7.61 (d, J=7.42 Hz, 2H), 7.78 (d, J=7.42 Hz,
2H). MS (ESI): found [M þ H]þ, 642.2.
N1,N3-Dimethyl-5-[4-[(2R,3S,4S,5S,6R)-3,4,5-trihydroxy-6-
(hydroxymethyl)tetrahydropyran-2-yl]oxyphenyl]benzene-1,3-
dicarboxamide (15b). 15a (0.050 g, 0.11 mmol) was stirred with
15 mL of MeNH2/EtOH (33 wt %) at room temperature for
40 h. The solvent was removed and the residue was dried in
vacuo to afford pure 15b (0.050 g) in quantitative yield. 1H
NMR (300 MHz, methanol-d4) δ ppm 2.96 (s, 6 H), 3.61-3.66
(m, 1 H), 3.67-3.84 (m, 3 H), 3.86-3.97 (m, 1 H), 4.04 (dd, J=
3.30, 1.92 Hz, 1 H), 5.56 (d, J = 1.92 Hz, 1 H), 7.21-7.34 (m,
2 H), 7.63-7.74 (m, 2 H), 8.13-8.26 (m, 3 H). MS (ESI): found
[M þ H]þ, 447.4.
N-Methyl-3-[4-[(2R,3S,4S,5S,6R)-3,4,5-trihydroxy-6-(hydro-
xymethyl)tetrahydropyran-2-yl]oxyphenyl]benzamide (14a). 14a
was prepared by direct amination of 8e, using the same method
1
as for 15b, in quantitative yield and in pure form. H NMR
(2S,3S,4S,5S,6R)-2-(4-Aminobutoxy)-6-(hydroxymethyl)-
tetrahydropyran-3,4,5-triol (19). 18 (0.106 g, 0.165 mmol) was
stirred in 6 mL of methanol with a catalytic amount of sodium
methoxide (0.02M) at room temperature overnight. The solvent
was removed and the residue was purified by silica gel chroma-
tography with a combination of aqueous ammonia and methanol
as eluent to afford 19 (0.036 g) in 88% yield. 1H NMR (300 MHz,
CD3OD) δ ppm 1.46-1.78 (m, 4H), 2.61-2.88 (m, 2H),
3.41-3.55 (m, 2H), 3.59 (t, J= 9.3 Hz, 1H), 3.65-3.87 (m, 5H),
4.75 (d, J=1.65 Hz, 1H). MS (ESI): found [M þ H]þ, 252.1.
5(6)-FAM-mannoside (20). A mixture of 19 (0.018 g, 0.070
mmol), 5-(and -6)-carboxyfluorescein succinimidyl ester (0.022 g,
0.046 mmol) and triethylamine (0.058 g, 0.57 mmol) in DMF
(2 mL) was stirred overnight. After removal of the solvent, the
residue was purified by silica gel chromatography with dichlor-
omethane/methanol combination as eluent, giving rise to 20
(0.023 g) in 82% yield. MS (ESI): found [M þ H]þ, 610.6.
3-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]-N-[2-[2-[2-[[3-[4-[(2R,3S,4S,5S,
6R)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydropyran-2-yl]-
oxyphenyl]benzoyl]amino]ethoxy]ethoxy]ethyl]benzamide (22).
Under nitrogen atmosphere, at 0 °C anhydrous DMF (5 mL)
was added into the RB flask containing 14c (0.062 g, 0.165 mmol)
and HATU (0.069 g, 0.182 mmol). After the mixture was stirred
for 10 min, 1,2-bis(2-aminoethoxy)ethane (0.0123 g, 0.083 mmol)
and then N,N-diisopropylethylamine (0.024 g, 0.182 mmol)
were added. The mixture was stirred overnight while being
(300 MHz, methanol-d4) δ ppm 2.94 (s, 3 H), 3.63 (ddd, J=7.28,
4.81, 2.47 Hz, 1 H), 3.69-3.85 (m, 3 H), 3.87-3.98 (m, 1 H), 4.04
(dd, J=3.30, 1.65 Hz, 1 H), 5.54 (d, J=1.92 Hz, 1 H), 7.10-7.27
(m, 2 H), 7.42-7.53 (m, 1 H), 7.54-7.65 (m, 2 H), 7.66-7.79 (m,
2 H), 8.01 (t, J = 1.79 Hz, 1 H). MS (ESI): found [M þ H]þ,
390.1.
5-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzene-1,3-dicarboxylic Acid
(15c). 15a (0.025 g, 0.056 mmol) was added into 7 mL of metha-
nol. Then 0.20 M NaOH aqueous (3 mL) was added. The
mixture was stirred at room temperature overnight. Hþ ex-
change resin (DOWEX 50WX4-100) was added to neutralize the
mixture. The resin was filtered off and the filtrate was concen-
trated and then dried in vacuo, giving rise to pure product 15c
(0.023 g) in quantitative yield. 1H NMR (300 MHz, methanol-
d4) δ ppm 3.59-3.67 (m, 1 H), 3.70-3.84 (m, 3 H), 3.94 (dd, J=
3.30, 9.30 Hz, 1 H), 4.05 (dd, J=3.30, 1.80 Hz, 1 H), 5.56 (d, J=
1.80 Hz, 1 H), 7.26 (m, 2 H), 7.63 (m, 2 H), 8.41 (d, J=1.50 Hz,
2 H), 8.57 (t, J = 1.50 Hz, 1 H). MS (ESI): found [M þ Na]þ,
443.0.
3-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]benzoic Acid (14c). 14c was pre-
pared through the hydrolysis of 8e, using the same method as for
15c, in quantitative yield and in pure form. 1H NMR (300 MHz,
methanol-d4) δ ppm 3.50-3.60 (m, 1 H), 3.61-3.77 (m, 3 H),
3.81-3.91 (m, 1 H), 3.96 (dd, J=3.16, 1.79 Hz, 1 H), 5.47 (d, J=

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12 4791
warmed to room temperature naturally. The solvent was re-
moved and the residue was purified by HPLC (C18, 15 mmꢀ150
mm column; eluent acetonitrile/water (0.1% TFA)) to give 22
(0.044 g) in 82% yield. 1H NMR (300 MHz, methanol-d4) δ ppm
3.51-3.65 (m, 6 H), 3.65-3.82 (m, 14 H), 3.88-3.96 (m, 2 H),
4.03 (dd, J = 3.57, 1.92 Hz, 2 H), 5.52 (d, J = 1.65 Hz, 2 H),
7.14-7.24 (m, 4 H), 7.40-7.50 (m, 2 H), 7.53-7.63 (m, 4 H),
7.66-7.77 (m, 4 H), 8.00 (t, J=1.65 Hz, 2 H). MS (ESI): found
[M þ H]þ, 865.9.
coordinated by Asp-47, the imidazole group of His-45, and
water molecules that are not well resolved. All copies of FimH
and compound 8e are essentially the same except for differences
in the orientation of the 8e carboxymethyl group, which in all
˚
cases is within hydrogen-bonding distance (2.8-3.0 A) of the
R98/D50 salt bridge. An initial map calculated with Fo - Fc
coefficients showed unambiguous difference density for 8e
occupying the extended mannose binding pocket of all four
copies of FimH present in the asymmetric unit. Subsequent
TLS refinement was performed with REFMAC16 using chemi-
cal restraints for 8e generated by eLBOW within the PHENIX
suite.17
3-[4-[(2R,3S,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)-
tetrahydropyran-2-yl]oxyphenyl]-N-[2-[2-[[3-[4-[(2R,3S,4S,5S,6R)-
3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydropyran-2-yl]oxyp-
henyl]benzoyl]amino]ethoxy]ethyl]benzamide (21). 21 was pre-
pared using the same method as for 22. Yield: 76%): 1H NMR
(300 MHz, methanol-d4) δ ppm 3.58-3.66 (m, 6 H), 3.68-3.82
(m, 10 H), 3.93 (dd, J=9.30, 3.30 Hz, 2 H), 4.03 (dd, J=3.60,
2.10 Hz, 2 H), 5.53 (d, J = 1.50 Hz, 2 H), 7.14-7.22 (m, 4 H),
7.36-7.42 (m, 2 H), 7.52-7.58 (m, 4 H), 7.64-7.72 (m, 4 H),
7.99 (t, J=1.50 Hz, 2 H). MS (ESI): found [M þ H]þ, 821.5.
FimH Fluorescence Polariazation/Anisotropy Assay. Each
compound tested was prepared as a 10 mM stock solution in
anhydrous DMSO (Sigma, St. Louis, MO) in batches of seven
test compounds (plus ButylRman as a positive control) on a 96-
well microtiter plate (Evergreen Plastics). Each compound was
prepared as a series of 2-fold serial dilutions by the Sciclone
robotic liquid handler (Caliper life sciences, Hopkinton, MA)
with 11 final compound concentrations ranging from 0 to 10 μM.
One microliter of prepared compound was then dry-transferred to
each of two 96-well microtiter plates (Corning) which were then
filled with 100 μL of assay mixture (200 nM FimH, 20 nM 5(6)-
FAM-mannoside (20), 50 mM Tris, 100 mM NaCl, 1 mM EDTA,
5 μg BSA, and 5 mM BME). Plates were assayed by reading fluo-
rescence anisotropy using the BioTek Synergy-2 microplate reader
(Winooski, VT), taking 100 readings/well using a stabilized tungs-
ten lamp followed by data averaging producing 1 data point per
well. Sigma Plot was used for fitting data to a one-site competition
binding curve (FP vs compound concentration) to generate an
EC50 for each compound.
Acknowledgment. We appreciate the assistance of Jay Nix
at beamline 4.2.2 of the Advanced Light Source with expert
data collection. This work was supported in part by the
National Institutes of Health (NIH) Grants 1RC1DK086378
and AI048689.
Supporting Information Available: 1H NMR spectra of final
compounds and intermediates, FP assay graphical result mea-
suring the KD of 5(6)-FAM (20), and details of the X-ray data
collection and refinement of FimH-8e. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
References
(1) Waksman, G.; Hultgren, S. J. Structural biology of the chaperone-
usher pathway of pilus biogenesis. Nat. Rev. Microbiol. 2009, 7,
765–774.
(2) (a) Rosen, D. A.; Pinkner, J. S.; Walker, J. N.; Elam, J. S.; Jones,
J. M.; Hultgren, S. J. Molecular variations in Klebsiella pneumoniae
and Escherichia coli FimH affect function and pathogenesis in the
urinary tract. Infect. Immun. 2008, 76, 3346–3356. (b) Chen, S. L.;
Hung, C. S.; Pinkner, J. S.; Walker, J. N.; Cusumano, C. K.; Li, Z.;
Bouckaert, J.; Gordon, J. I.; Hultgren, S. J. Positive selection identifies
an in vivo role for FimH during urinary tract infection in addition to
mannose binding. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 22439–
22444. (c) Abraham, S. N.; Sun, D.; Dale, J. B.; Beachey, E. H.
Conservation of the D-mannose-adhesion protein among type 1 fim-
briated members of the family Enterobacteriaceae. Nature 1988, 336,
682–684. (d) Martinez, J. J.; Mulvey, M. A.; Schilling, J. D.; Pinkner,
J. S.; Hultgren, S. J. Type 1 pilus-mediated bacterial invasion of bladder
epithelial cells. EMBO J. 2000, 19, 2803–2812. (e) Mulvey, M. A.;
Lopez-Boado, Y. S.; Wilson, C. L.; Roth, R.; Parks, W. C.; Heuser, J.;
Hultgren, S. J. Induction and evasion of host defenses by type 1-piliated
uropathogenic Escherichia coli. Science 1998, 282, 1494–1497. (f) Wright,
K. J.; Seed, P. C.; Hultgren, S. J. Development of intracellular bacterial
communities of uropathogenic Escherichia coli depends on type 1 pili.
Cell. Microbiol. 2007, 9, 2230–2241. (g) Hung, C. S.; Bouckaert, J.;
Hung, D.; Pinkner, J.; Widberg, C.; DeFusco, A.; Auguste, C. G.;
Strouse, R.; Langermann, S.; Waksman, G.; Hultgren, S. J. Structural
basis of tropism of Escherichia coli to the bladder during urinary tract
infection. Mol. Microbiol. 2002, 44, 903–915.
(3) (a) Mulvey, M. A.; Schilling, J. D.; Hultgren, S. J. Establishment of
a persistent Escherichia coli reservoir during the acute phase of a
bladder infection. Infect. Immun. 2001, 69, 4572–4579. (b) Ander-
son, G. G.; Dodson, K. W.; Hooton, T. M.; Hultgren, S. J. Intracellular
bacterial communities of uropathogenic Escherichia coli in urinary tract
pathogenesis. Trends Microbiol. 2004, 12, 424–430. (c) Anderson,
G. C.; Palermo, J. J.; Schilling, J. D.; Roth, R.; Heuser, J.; Hultgren, S. J.
Intracellular bacterial biofilm-like pods in urinary tract infections.
Science 2003, 301, 105. (d) Justice, S. S.; Hung, C.; Theriot, J. A.;
Fletcher, D. A.; Anderson, G. G.; Footer, M. J.; Hultgren, S. J. Dif-
ferentiation and developmental pathways of uropathogenic Escherichia
coli in urinary tract pathogenesis. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 1333–1338.
X-ray Crystallography. The adhesin domain of FimH (here-
after “FimH”) (residues 1-175 with a C-terminal 6-histidine
tag) was cloned into a pTRC99 plasmid and expressed in C600
E. coli. FimH was purified from bacterial periplasm by passage
over cobalt affinity (Talon, Clontech) and Q Sepharose (GE
Healthcare) columns. Protein was subsequently dialyzed against
10 mM MES at pH 6.5, concentrated to 15 mg/mL, and incu-
bated for 30 min with 2 mM compound 8e prior to cocrystal-
lization. FimH-8e crystallized in 20% ethanol, 100 mM imida-
zole, pH 8.0, and 200 mM MgCl2. Tetragonal bipyramidal
crystals measuring 150 μm ꢀ 150 μm formed in 2 days and did
not diffract unless slowly dehydrated over the course of at least
12 h. Crystals of the FimH-8e complex were therefore treated
by soaking individual crystals in a stabilizing solution initially
containing 15% ethanol, 100 mM imidazole, pH 8.0, 200 mM
MgCl2, 10% PEG200, 5% glycerol, and 1 mM compound 8e.
Individual crystals were placed in 150 μL of stabilizing solution
in open spot plates and allowed to dehydrate to one-third their
original volume over the course of 24 h. Crystals were then
harvested without further cryopretection and plunged into
˚
liquid nitrogen. Crystals were diffracted to 2.9 A at beamline 4.2.2
at the Advanced Light Source. Data collection and refinement
statistics are summarized in Table 1 of the Supporting Information.
The structure of FimH-8e was solved by molecular replace-
ment with the program PHASER15 using a previously solved
model of FimH with its R-D-methylmannoside (MethylRman)
ligand removed as the search model (PDB code 1UWF6a). There
are four FimH-8e complexes in the asymmetric unit. Each
model of FimH is represented by residues 1-158; further
C-terminal residues and the 6-His tag are absent and presumed
to be disordered. Each copy of FimH is bound to a single cation,
presumed to be calcium from the crystallization condition that is
(4) Justice, S. S.; Hunstad, D. A.; Seed, P. C.; Hultgren, S. J. Matura-
tion of intracellular escherichia coli communities requires SurA.
Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 19884–19889.
(5) (a) Langermann, S.; Palaszynski, S.; Barnhart, M.; Auguste, G.;
Pinkner, J. S.; Burlein, J.; Barren, P.; Koenig, S.; Leath, S.; Jones,
C. H.; Hultgren, S. J. Prevention of mucosal escherichia coli
infection by FimH-adhesin-based systemic vaccination. Science
1997, 276, 607–611. (b) Cegelski, L.; Pinkner, J. S.; Hammer, N. D.;
Cusumano, C. K.; Hung, C. S.; Chorell, E.; Aberg, V.; Walker,
J. N.; Seed, P. C.; Almqvist, F.; Chapman, M. R.; Hultgren, S. J.

4792 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 12
Han et al.
Small-molecule inhibitors target Escherichia coli amyloid biogenesis
and biofilm formation. Nat. Chem. Biol. 2009, 5, 913–919.
(6) (a) Bouckaert, J.; Berglund, J.; Schembri, M.; de Genst, E.; Cools,
(7) Firon, N.; Ashkenazi, S; Mirelman, D.; Ofek, I.; Sharon, N.
Aromatic alpha-glycosides of mannose are powerful inhibitors of
the adherence of type 1 fimbriated Escherichia coli to yeast and
intestinal epithelial cells. Infect. Immun. 1987, 55, 472–476.
(8) Meuwly, R.; Vasella, A. Synthesis of 1,2-cis-configurated glyco-
sylphosphonates. Helv. Chim. Acta 1986, 69, 25–34.
(9) (a) Hultgren, S. J.; Duncan, J. L.; Schaeffer, A. J.; Amundsen, S. K.
Mannose-sensitive haemagglutination in the absence of piliation in
Escherichia coli. Mol. Microbiol. 1990, 4, 1311–1318. (b) Slonim,
L. N.; Pinkner, J. S.; Branden, C. I.; Hultgren, S. J. Interactive surface in
the PapD chaperone cleft is conserved in pilus chaperone superfamily
and essential in subunit recognition and assembly. EMBO J. 1992, 11
(13), 4747–56.
€
L.; Wuhrer, M.; Hung, C.-S.; Pinkner, J.; Slattegard, R.; Zavialov,
˚
A.; Choudhury, D.; Langermann, S.; Hultgren, S. J.; Wyns, L.;
Klemm, P.; Oscarson, S.; Knight, S. D.; de Greve, H. Receptor
binding studies disclose a novel class of high-affinity inhibitors of
the Escherichia coli FimH adhesion. Mol. Microbiol. 2005, 55, 441–
455. (b) Gouin, S. G.; Wellens, A.; Bouckaert, J.; Kovensky, J.
Synthetic multimeric heptyl mannosides as potent antiadhesives of
uropathogenic Escherichia coli. ChemMedChem 2009, 4, 749–755.
€
(c) Wellens, A.; Garofalo, C.; Nguyen, H.; Van Gerven, N.; Slattegård,
R.; Hernalsteens, J.-P.; Wyns, L.; Oscarson, S.; De Greve, H.; Hultgren,
S.; Bouckaert, J. Intervening with urinary tract infections using anti-
adhesives based on the crystal structure of the FimH-oligomannose-3
complex. PLoS One 2008, 3, No. e2040. (d) Touaibia, M.; Wellens, A.;
Shiao, T. C.; Wang, Q.; Sirois, S.; Bouckaert, J.; Roy, R. Mannosylated
G(0) dendrimers with nanomolar affinities to Escherichia coli FimH.
ChemMedChem 2007, 2, 1190–1201. (e) Touaibia, M.; Shiao, T. C.;
Papadopoulos, A.; Vaucher, J.; Wang, Q. G.; Benhamioud, K.; Roy, R.
Tri- and hexavalent mannoside clusters as potential inhibitors of type 1
fimbriated bacteria using pentaerythritol and triazole linkages. Chem.
Commun. 2007, 380–382. (f) Touaibia, M.; Roy, R. Glycodendrimers
as anti-adhesion drugs against type 1 fimbriated E. coli uropathogenic
infections. Mini-Rev. Med. Chem. 2007, 7, 1270–1283. (g) Nagahori,
N.; Lee, R. T.; Nishimura, S.; Page, D.; Roy, R.; Lee, Y. C. Inhibition of
adhesion of type 1 fimbriated Escherichia coli to highly mannosylated
ligands. ChemBioChem 2002, 3, 836–844. (h) Furneaux, R. H.; Pakulski,
Z.; Tyler, P. C. New mannotriosides and trimannosides as potential
ligands for mannose-specific binding proteins. Can. J. Chem. 2002, 80,
964–972. (i) Lindhorst, T. K.; Kieburg, C.; Krallmann-Wenzel, U.
Inhibition of the type 1 fimbriae-mediated adhesion of Escherichia coli
to erythrocytes by multiantennary alpha-mannosyl clusters: the effect
of multivalency. Glycoconjugate J. 1998, 15, 605–613. (j) Kotter, S.;
Krallmann-Wenzel, U.; Ehlers, S.; Lindhorst, T. K. Multivalent ligands
for the mannose-specific lectin on type 1 fimbriae of Escherichia coli:
syntheses and testing of trivalent alpha-D-mannoside clusters. J. Chem.
Soc., Perkin Trans. 1 1998, 2193–2200. (k) Sperling, O.; Fuchs, A.;
Lindhorst, T. K. Evaluation of the carbohydrate recognition domain of
the bacterial adhesin FimH: design, synthesis and binding properties of
mannoside ligands. Org. Biomol. Chem. 2006, 4, 3913–3922.
(10) Baker, N. A.; Sept, D.; Joseph, S; Holst, M. J.; McCammon, J. A.
Electrostatics of nanosystems: application to microtubules and the
ribosome. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 10037–10041.
(11) Topliss, J. G. Utilization of operational schemes for analog synth-
esis in drug design. J. Med. Chem. 1972, 15 (10), 1006–1011.
(12) (a) Cusumano, C. K.; Hultgren, S. J. Bacterial adhesion. A source
of alternate antibiotic targets. IDrugs 2009, 12, 699–705. (b) Cegelski,
L.; Marshall, G. R.; Eldridge, G. R.; Hultgren, S. J. The biology and
future prospects of antivirulence therapies. Nat. Rev. Microbiol. 2008,
6 (1), 17–27.
(13) Ernst, B.; Magnani, J. L. From carbohydrate leads to glycomimetic
drugs. Nat. Rev. Drug Discovery 2009, 8, 661–677.
(14) Kogan, T. P.; Dupre, B.; Bui, H.; McAbee, K. L.; Kassir, J. M.;
Scott, I. L.; Hu, X.; Vanderslice, P.; Beck, P. J.; Dixon, R. A. F.
Novel synthetic inhibitors of selectin-mediated cell adhesion:
synthesis of 1,6-bis[3-(3-carboxymethylphenyl)-4-(2-R-D-manno-
pyranosyloxy)phenyl]hexane (TBC1269). J. Med. Chem. 1998, 41,
1099–1111.
(15) McCoy, A. J.; Grosse-Kunstleve, R. W.; Adams, P. D.; Winn,
M. D.; Storoni, L. C.; Read, R. J. Phaser crystallographic software.
J. Appl. Crystallogr. 2007, 40, 658–674.
(16) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Refinement of
macromolecular structures by the maximum-likelihood method.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 1997, 53, 240–55.
(17) Moriarty, N. W.; Grosse-Kunstleve, R. W.; Adams, P. D. Electro-
nic ligand builder and optimization workbench (eLBOW): a tool
for ligand coordinate and restraint generation. Acta Crystallogr.,
Sect. D: Biol. Crystallogr. 2009, 65, 1074–1080.